The batteries we don’t need anymore

I continue on the thread I started to develop in my last update in French, titled ‘De quoi parler à la prochaine réunion de faculté’, i.e. I am using that blog, and the fact of writing, to put some order in the almost ritual mess that happens at the beginning of the academic year. New calls for tenders start in the ministerial grant programs, new syllabuses need to be prepared, new classes start. Ordinary stuff, mind you, this is just something about September, as if I were in Vivaldi’s ‘Four seasons’: the hot, tumultuous Summer slowly folds into the rich, textured, and yet implacably realistic Autumn.

My central idea is to use some of the science which I dove into during the summer holidays as an intellectual tool for putting order in that chaos. That almost new science of mine is mostly based on the theory of complex systems, and my basic claim is that technological change is an emergent phenomenon in complex social systems. We don’t know why exactly our technologies change the way they change. We can trace the current technologies back to their most immediate ancestors and sometimes we can predict their most immediate successors, but that’s about it. Futuristic visions of technologies that could be there in 50 years from now are already some kind of traditional entertainment. The concept of technological progress, when we try to find a developmental logic in the historically known technological change, is usually standing on wobbly legs, on the other hand. Yes, electricity allowed the emergence of medical technologies used in hospitals, and that saved a lot of human lives, but there is no way Thomas Edison could have known that. The most spectacular technological achievements of mankind, such as the Egyptian pyramids, the medieval cathedrals, the Dutch windmills from the 16th century, or the automobile, seen from the historical distance, look ambiguous. Yes, it all solved some problems, but it facilitated the emergence of new problems. The truly unequivocal benefit of those technological leaps, which could have been actually experienced by the people who made them, was to learn how to develop technologies.

The studies I did during the Summer holidays 2021 focused on four essential, mathematical models of emergent technological change: cellular automata, flock of birds AKA particle swarm, ants’ nest, and imperfect Markov chains. I start with passing in review the model of cellular automata. At any given moment, the social complexity can be divided into a finite number of social entities (agents). They can be individual humans, businesses, NGOs, governments, local markets etc. Each such entity has an immediate freedom of movement, i.e. a finite number of one-step moves. The concept is related to the theory of games and corresponds to what happens in real life. When we do something social, we seldom just rush forwards. Most frequently, we make one step, observe the outcomes, adjust, then we make the next step etc. When all social agents do it, the whole social complexity can be seen as a collection of cells, or pixels. Each such cell (pixel) is a local state of being in society. A social entity can move into that available state, or not, at their pleasure and leisure. All the one-step moves a social entity can make translate into a trajectory it can follow across the social space. Collective outcomes we strive for and achieve can be studied as temporary complex states of those entities following their respective trajectories. The epistemological trick here is that individual moves and their combinations can be known for sure only ex post. All we can do ex ante is to define the possible states, and just wait where does the reality go.

As we are talking about the possible states of social complexity, I found an interesting mathematical mindf**k at quite an unexpected source, namely in the book titled ‘Aware. The Science and Practice of Presence. The Groundbreaking Meditation Practice’ by Daniel J. Siegel [Penguin Random House LLC, 2018, Identifiers: LCCN 2018016987 (print), LCCN 2018027672 (ebook), ISBN 9780143111788, ISBN 9781101993040 (hardback)]. This is a mathematical way of thinking, apparently taken from quantum physics. Here is the essence of it. Everything that happens does so as 100% probability of the given thing happening. Each phenomenon which takes place is the actualization of the same phenomenon being just likely to happen.

Actualization of probability can be seen as collision of two vehicles in traffic. When the two vehicles are at a substantial distance from each other, the likelihood of them colliding is zero, for all practical purposes. As they converge towards each other, there comes a point when they become sort of provisionally entangled, e.g. they find themselves heading towards the same crossroads. The probability of collision increases slightly, and yet it is not even the probability of collision, it is just the probability that these two might find themselves in a vicinity conducive to a possible collision. Nothing to write home about, yet, like really. It can be seen as a plateau of probability slowly emerging out of the initial soup of all the things which can possibly happen.

As the two cars drive closer and closer to the crossroads in question, the panoply of possible states narrows down. There is a very clear chunk of reality which gains in likelihood, as if it was a mountain range pushing up from the provisional plateau. There comes a point where the two cars (and their drivers) just come on collision course, and there is no way around it, and this is a peak of 100% probability. Boom! Probability is being consumed.

What do those cars have in common with meditation and with the emergence of technological change? As regards meditation, thought can be viewed as a progressively emerging actualization of something that was just a weak probability, sort of a month ago it was just weakly probable that today I would think what I think, it became much more likely yesterday, as the thoughts from yesterday have an impact on the thoughts of today, and today it all comes to fruition, i.e. to the 100% probability. As regards emergent technological change, the way technology changes today can be viewed as actualization of something that was highly probable last year, just somehow probable 10 years ago, and had been just part of the amorphous soup of probability 30 years ago. Those trajectories followed by individual agents inside social complexity, as defined in the theory of cellular automata, are entangled together precisely according to that pattern of emergent probabilities. Two businesses coming up with two mutually independent, and yet similar technologies, are like two peak actualizations of 100% probability in a plateau of probable technological change, which, in turn, has been slowly emerging for some time.

Those other theories I use explain and allow to model mathematically that entanglement. The theory of particle swarm, pertinent to flocks of birds, assumes that autonomous social agents strive for a certain level of behavioural coupling. We expect some level of predictability from others, and we can cooperate with others when we are satisfactorily predictable in our actions. The strive for social coherence is, therefore, one mechanism of entanglement between individual trajectories of cellular automata. The theory of ants’ nest focuses on a specific category of communication systems in societies, working like pheromones. Ants organize by marking, reinforcing and following paths across their environment, and their pheromones serve as markers and reinforcement agents for those paths. In human societies, there are social pheromones. Money and financial markets make probably the most obvious example, but scientific publications are another one. The more scientific articles are being published on a given topic, the more likely are other articles being written on the same topic, until the whole thing reaches a point of saturation, when some ants (pardon me, scientists) start thinking about another path to mark with intellectual pheromones.

Cool. I have (OK, we have) complex social states, made of entangled probabilities that something specific happens, and they encompass technology. Those complex states change, i.e. one complex state morphs into another. Now, how the hell can I know, as a researcher, what is happening exactly? Such as the theory of complex systems suggests it, I can never know exactly, for one, and I need to observe, for two. As I don’t know exactly what is it exactly, that thing which I label ‘technological change’, it is problematic to set too many normative assumptions as for which specific path that technological change should take. I think this is the biggest point of contention as I apply my theory, such as I have just outlined it, to my main field of empirical research, namely energy economics, and technological change in the sector of energy. The more I do that research, the more convinced I am that the so-called ‘energy policies’, ‘climate policies’ etc. are politically driven bullshit based on wishful thinking, with not much of a chance to bring the positive change we expect. I have that deep feeling that setting a strategy for future innovations in our business/country/world is very much like that Polish expression ‘sharing the skin of a bear which is still running in the woods’. First, you need to kill the bear, only then you can bicker about who takes what part of the skin. In the case of innovation, long-term strategies in that domain consist in predicting what we will do when we have something we don’t even know yet what is it exactly.

I am trying to apply this general theory in the grant applications which I am in charge of preparing now, and in my teaching. We have that idea, at the faculty, to apply for funding to study the market of electric vehicles in Europe and in Poland. This is an interesting situation as regards business models. In the US, the market of electric cars is clearly divided among three categories of players. There is Tesla, which is a category and an industry in itself, with its peculiar strategy of extreme vertical integration. Then there are the big, classical car makers, such as Toyota, General Motors etc., with their business models based on rather a short vertical chain of value added inside the business, and a massive supply chain upstream of the house. Finally, there is a rising tide of small start-ups in the making of electric vehicles. I wonder what I could be in Europe. As our European market of electric vehicles is taking off, it is dominated by the incumbent big manufacturers, the old school ones, with Tesla building a factory in Germany, and progressively building a beachhead in the market. There is some timid movement towards small start-up businesses in the field, but it is really timid. In my home country, Poland, the most significant attempt at starting up an electric vehicle made in Poland is a big consortium of state-controlled companies, running under the name of ‘Electromobility Poland’.  

I have that intuition, which I provisionally express as a working hypothesis, namely that business models are an emergent property of technologies which they use. As regards the market of electric vehicles, it means that Tesla’s business model is not an accidental explosion of Elon Musk’s genius mind: it is an emergent characteristic of the technologies involved.

Good. I have some theory taking shape, nice and easy. I let it ripen a bit, and I start sniffing around for facts. What is a business model, in my mind? It is the way of operating the chain of value added, and getting paid for it, in the first place. Then, it is the way of using capital. I noticed that highly innovative environments force businesses to build up and keep large amounts of cash money, arguably to manage the diverse uncertainties emerging as technologies around morph like hell. In some cases, e.g. in biotech, the right business model for rapid innovation is a money-sucker, with apparently endless pay-ins of additional equity by the shareholders, and yet with a big value in terms of technological novelty created. I can associate that phenomenon of vacuum cleaning equity with the case of Tesla, who just recently started being profitable, and had gone through something like a decade in permanent operational loss. That is all pertinent to fixed costs, thus to the cash we need to build up and keep in place the organizational structure required for managing the value chain the way we want to manage it.

I am translating those loose remarks of mine into observable phenomena. Everything I have just mentioned is to be found in the annual financial reports. This is my first source of information. When I want to study business models in the market of electric vehicles, I need to look into financial and corporate reports of businesses active in the market. I need to look into the financial reports of Mercedes Benz, BMW, Renault, PSA, Volkswagen, Fiat, Volvo, and Opel – thus the European automotive makers – and see how it is going, and whether whatever is going on can be correlated with changes in the European market of electric vehicles. Then, it is useful to look into the financial reports of global players present in the European market, e.g. Tesla, Toyota, Honda and whatnot, just to see what changes in them as the European market of electric vehicles is changing.

If my intuition is correct, i.e. if business models are truly an emergent property of technologies used, the fact of engaging into the business of electric vehicles should be correlated with some sort of recurrent pattern in those companies.         

Good. This is about the big boys in the playground. Now, I turn toward the small ones, the start-up businesses. As I already said, it is not like we have a crowd of them in the European industry of electric vehicles. The intuitive axis of research which comes to my mind is to look at start-ups active in the U.S., study their business models, and see if there is any chance of something similar emerging in Europe. Somehow tangentially to that, I think it would be interesting to check whether the plan of Polish government regarding ‘Electromobility Poland’, that is the plan to develop it with public and semi-public money, and then sell it to private investors, has any grounds and under what conditions it can be a workable plan.

Good. I have rummaged a bit in my own mind, time to do the same to other people. I mean, I am passing to reviewing the literature. I type ‘electric vehicles Europe business model’ at the https://www.sciencedirect.com/ platform, and I look at what’s popping up. Here comes the paper by Pardo-Bosch, F., Pujadas, P., Morton, C., & Cervera, C. (2021). Sustainable deployment of an electric vehicle public charging infrastructure network from a city business model perspective. Sustainable Cities and Society, 71, 102957., https://doi.org/10.1016/j.scs.2021.102957 . The abstract says: ‘The unprecedented growth of global cities together with increased population mobility and a heightened concern regarding climate change and energy independence have increased interest in electric vehicles (EVs) as one means to address these challenges. The development of a public charging infrastructure network is a key element for promoting EVs, and with them reducing greenhouse gas emissions attributable to the operation of conventional cars and improving the local environment through reductions in air pollution. This paper discusses the effectiveness, efficiency, and feasibility of city strategic plans for establishing a public charging infrastructure network to encourage the uptake and use of EVs. A holistic analysis based on the Value Creation Ecosystem (VCE) and the City Model Canvas (CMC) is used to visualise how such plans may offer public value with a long-term and sustainable approach. The charging infrastructure network implementation strategy of two major European cities, Nantes (France) and Hamburg (Germany), are analysed and the results indicate the need to involve a wide range of public and private stakeholders in the metropolitan areas. Additionally, relevant, and fundamental patterns and recommendations are provided, which may help other public managers effectively implement this service and scale-up its use and business model.

Well, I see there is a lot of work to do, as I read that abstract. I rarely find a paper where I have so much to argue with, just after having read the abstract. First of all, ‘the unprecedented growth of global cities’ thing. Actually, if you care to have a look at the World Bank data on urban land (https://data.worldbank.org/indicator/AG.LND.TOTL.UR.K2 ), as well as that on urban population (https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS ), you will see that urbanization is an ambiguous phenomenon, strongly region-specific. The central thing is that cities become increasingly distinct from the countryside, as types of human settlements. The connection between electric vehicles and cities is partly clear, but just partly. Cities are the most obvious place to start with EVs, because of the relatively short distance to travel between charging points. Still, moving EVs outside the cities, and making them functional in rural areas, is the next big challenge.

Then comes the ‘The development of a public charging infrastructure network is a key element for promoting EVs’ part. As I studied the thing in Europe, the network of charging stations, as compared to the fleet of EVs in the streets is so dense that we have like 12 vehicles per charging station on average, across the European Union. There is no way a private investor can have it for their money, when financing a private charging station, with that average density. We face a paradox: there are so many publicly funded charging stations, in relation to the car fleet out there, that private investment gets discouraged. I agree that it could be an acceptable transitory state in the market, although it begs the question whether private charging stations are a viable business in Europe. Tesla has based a large part of its business model in the US precisely on the development of their own charging stations. Is it a viable solution in Europe?

Here comes another general remark, contingent to my hypothesis of business models being emergent on the basis of technologies. Automotive technologies in general, thus the technology of a vehicle moving by itself, regardless the method of propulsion (i.e. internal combustion vs electric) is a combination of two component technologies. Said method of propulsion is one of them, and the other one is the technology of distributing the power source across space. Electric vehicles can be viewed as cousins to tramways and electric trains, with just more pronounced a taste for independence: instead of drinking electricity from a permanent wiring, EVs carry their electricity around with them, in batteries.

As we talk about batteries, here comes another paper in my cursory rummaging across other people’s science: Albertsen, L., Richter, J. L., Peck, P., Dalhammar, C., & Plepys, A. (2021). Circular business models for electric vehicle lithium-ion batteries: An analysis of current practices of vehicle manufacturers and policies in the EU. Resources, Conservation and Recycling, 172, 105658., https://doi.org/10.1016/j.resconrec.2021.105658 . Yes, indeed, the advent of electric vehicles creates a problem to solve, namely what to do with all those batteries. I mean two categories of batteries. Those which we need, and hope to acquire easily when the time comes for changing them in our vehicles, in the first place, and those we don’t need anymore and expect someone to take care of them swiftly and elegantly.       

I have proven myself wrong

I keep working on a proof-of-concept paper for the idea I baptized ‘Energy Ponds’. You can consult two previous updates, namely ‘We keep going until we observe’ and ‘Ça semble expérimenter toujours’ to keep track of the intellectual drift I am taking. This time, I am focusing on the end of the technological pipeline, namely on the battery-powered charging station for electric cars. First, I want to make myself an idea of the market for charging.

I take the case of France. In December 2020, they had a total of 119 737 electric vehicles officially registered (matriculated), which made + 135% as compared to December 2019[1]. That number pertains only to 100% electrical ones, with plug-in hybrids left aside for the moment. When plug-in hybrids enter the game, France had, in December 2020, 470 295 vehicles that need or might need the services of charging stations. According to the same source, there were 28 928 charging stations in France at the time, which makes 13 EVs per charging station. That coefficient is presented for 4 other European countries: Norway (23 EVs per charging station), UK (12), Germany (9), and Netherlands (4).

I look up into other sources. According to Reuters[2], there was 250 000 charging stations in Europe by September 2020, as compared to 34 000 in 2014. That means an average increase by 36 000 a year. I find a different estimation with Statista[3]: 2010 – 3 201; 2011 – 7 018; 2012 – 17 498; 2013 – 28 824; 2014 – 40 910; 2015 – 67 064; 2016 – 98 669; 2017 – 136 059; 2018 – 153 841; 2019 – 211 438; 2020 – 285 796.

On the other hand, the European Alternative Fuels Observatory supplies their own data at https://www.eafo.eu/electric-vehicle-charging-infrastructure, as regards European Union.

Number of EVs per charging station (source: European Alternative Fuels Observatory):

EVs per charging station
201014
20116
20123
20134
20145
20155
20165
20175
20186
20197
20209

The same EAFO site gives their own estimation as regards the number of charging stations in Europe:

Number of charging stations in Europe (source: European Alternative Fuels Observatory):

High-power recharging points (more than 22 kW) in EUNormal charging stations in EUTotal charging stations
201225710 25010 507
201375117 09317 844
20141 47424 91726 391
20153 39644 78648 182
20165 19070 01275 202
20178 72397 287106 010
201811 138107 446118 584
201915 136148 880164 016
202024 987199 250224 237

Two conclusions jump to the eye. Firstly, there is just a very approximate count of charging stations. Numbers differ substantially from source to source. I can just guess that one of the reasons for that discrepancy is the distinction between officially issued permits to build charging points, on the one hand, and the actually active charging points, on the other hand. In Europe, building charging points for electric vehicles has become sort of a virtue, which governments at all levels like signaling. I guess there is some boasting and chest-puffing in the numbers those individual countries report.  

Secondly, high-power stations, charging with direct current, with a power of at least 22 kWh,  gain in importance. In 2012, that category made 2,45% of the total charging network in Europe, and in 2020 that share climbed to 11,14%. This is an important piece of information as regards the proof-of-concept which I am building up for my idea of Energy Ponds. The charging station I placed at the end of the pipeline in the concept of Energy Ponds, and which is supposed to earn a living for all the technologies and installations upstream of it, is supposed to be powered from a power storage facility. That means direct current, and most likely, high power.   

On the whole, the www.eafo.eu site seems somehow more credible that Statista, with all the due respect for the latter, and thus I am reporting some data they present on the fleet of EVs in Europe. Here it comes, in a few consecutive tables below:

Passenger EVs in Europe (source: European Alternative Fuels Observatory):

BEV (pure electric)PHEV (plug-in-hybrid)Total
20084 1554 155
20094 8414 841
20105 7855 785
201113 39516313 558
201225 8913 71229 603
201345 66232 47478 136
201475 47956 745132 224
2015119 618125 770245 388
2016165 137189 153354 290
2017245 347254 473499 820
2018376 398349 616726 014
2019615 878479 7061 095 584
20201 125 485967 7212 093 206

Light Commercial EVs in Europe (source: European Alternative Fuels Observatory):

BEV (pure electric)PHEV (plug-in-hybrid)Total
2008253253
2009254254
2010309309
20117 6697 669
20129 5279 527
201313 66913 669
201410 04910 049
201528 61028 610
201640 926140 927
201752 026152 027
201876 286176 287
201997 36311797 480
2020120 7111 054121 765

Bus EVs in Europe (source: European Alternative Fuels Observatory):

BEV (pure electric)PHEV (plug-in-hybrid)Total
20082727
20091212
2010123123
2011128128
2012286286
2013376376
201438940429
2015420145565
2016686304990
20178884451 333
20181 6084862 094
20193 6365254 161
20205 3115505 861

Truck EVs in Europe (source: European Alternative Fuels Observatory):

BEV (pure electric)PHEV (plug-in-hybrid)Total
200855
200955
201066
201177
201288
20134747
20145858
20157171
201611339152
2017544094
201822240262
201959538633
2020983291 012

Structure of EV fleet in Europe as regards the types of vehicles (source: European Alternative Fuels Observatory):

Passenger EVLight commercial EVBus EVTruck EV
200893,58%5,70%0,61%0,11%
200994,70%4,97%0,23%0,10%
201092,96%4,97%1,98%0,10%
201163,47%35,90%0,60%0,03%
201275,09%24,17%0,73%0,02%
201384,72%14,82%0,41%0,05%
201492,62%7,04%0,30%0,04%
201589,35%10,42%0,21%0,03%
201689,39%10,33%0,25%0,04%
201790,34%9,40%0,24%0,02%
201890,23%9,48%0,26%0,03%
201991,46%8,14%0,35%0,05%
202094,21%5,48%0,26%0,05%

Summing it up a bit. The market of Electric Vehicles in Europe seems being durably dominated by passenger cars. There is some fleet in other categories of vehicles, and there is even some increase, but, for the moment, in all looks more like an experiment. Well, maybe electric buses turn up sort of more systemically.

The proportion between the fleet of electric vehicles and the infrastructure of charging stations still seems to be in the phase of adjustment in the latter to the abundance of the former. Generally, the number of charging stations seems to be growing slower than the fleet of EVs. Thus, for my own concept, I assume that the coefficient of 9 EVs per charging station, on average, will stand still or will slightly increase. For the moment, I take 9. I assume that my charging stations will have like 9 habitual customers, plus a fringe of incidental ones.

From there, I think in the following terms. The number of times the average customer charges their car depends on the distance they cover. Apparently, there is like a 100 km  50 kWh equivalence. I did not find detailed statistics as regards distances covered by electric vehicles as such, however I came by some Eurostat data on distances covered by all passenger vehicles taken together: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Passenger_mobility_statistics#Distance_covered . There is a lot of discrepancy between the 11 European countries studied for that metric, but the average is 12,49 km per day. My average 9 customers would do, in total, an average of 410,27 of 50 kWh charging purchases per year. I checked the prices of fast charging with direct current: 2,3 PLN per 1 kWh in Poland[4],  €0,22 per 1 kWh in France[5], $0,13 per 1 kWh in US[6], 0,25 pence per 1 kWh in UK[7]. Once converted to US$, it gives $0,59 in Poland, $0,26 in France, $0,35 in UK, and, of course, $0,13 in US. Even at the highest price, namely that in Poland, those 410,27 charging stops give barely more than $12 000 a year.

If I want to have a station able to charge 2 EVs at the same time, fast charging, and counting 350 kW per charging pile (McKinsey 2018[8]), I need 700 kW it total. Investment in batteries is like $600 ÷ $800 per 1 kW (Cole & Frazier 2019[9]; Cole, Frazier, Augustine 2021[10]), thus 700 * ($600 ÷ $800) = $420 000 ÷ $560 000. There is no way that investment pays back with $12 000 a year in revenue, and I haven’t even started talking about paying off on investment in all the remaining infrastructure of Energy Ponds: ram pumps, elevated tanks, semi-artificial wetlands, and hydroelectric turbines.

Now, I revert my thinking. Investment in the range of $420 000 ÷ $560 000, in the charging station and its batteries, gives a middle-of-the-interval value of $490 000. I found a paper by Zhang et al. (2018[11]) who claim that a charging station has chances to pay off, as a business, when it sells some 5 000 000 kWh a year. When I put it back-to-back with the [50 kWh / 100 km] coefficient, it gives 10 000 000 km. Divided by the average annual distance covered by European drivers, thus by 4 558,55 km, it gives 2 193,68 customers per year, or some 6 charging stops per day. That seems hardly feasible with 9 customers. I assume that one customer would charge their electric vehicle no more than twice a week, and 6 chargings a day make 6*7 = 42 chargings, and therefore 21 customers.

I need to stop and think. Essentially, I have proven myself wrong. I had been assuming that putting a charging station for electric vehicles at the end of the internal value chain in the overall infrastructure of Energy Ponds will solve the problem of making money on selling electricity. Turns out it makes even more problems. I need time to wrap my mind around it.


[1] http://www.avere-france.org/Uploads/Documents/161011498173a9d7b7d55aef7bdda9008a7e50cb38-barometre-des-immatriculations-decembre-2020(9).pdf

[2] https://www.reuters.com/article/us-eu-autos-electric-charging-idUSKBN2C023C

[3] https://www.statista.com/statistics/955443/number-of-electric-vehicle-charging-stations-in-europe/

[4] https://elo.city/news/ile-kosztuje-ladowanie-samochodu-elektrycznego

[5] https://particulier.edf.fr/fr/accueil/guide-energie/electricite/cout-recharge-voiture-electrique.html

[6] https://afdc.energy.gov/fuels/electricity_charging_home.html

[7] https://pod-point.com/guides/driver/cost-of-charging-electric-car

[8] McKinsey Center for Future Mobility, How Battery Storage Can Help Charge the Electric-Vehicle Market?, February 2018,

[9] Cole, Wesley, and A. Will Frazier. 2019. Cost Projections for Utility-Scale Battery Storage.

Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-73222. https://www.nrel.gov/docs/fy19osti/73222.pdf

[10] Cole, Wesley, A. Will Frazier, and Chad Augustine. 2021. Cost Projections for UtilityScale Battery Storage: 2021 Update. Golden, CO: National Renewable Energy

Laboratory. NREL/TP-6A20-79236. https://www.nrel.gov/docs/fy21osti/79236.pdf.

[11] Zhang, J., Liu, C., Yuan, R., Li, T., Li, K., Li, B., … & Jiang, Z. (2019). Design scheme for fast charging station for electric vehicles with distributed photovoltaic power generation. Global Energy Interconnection, 2(2), 150-159. https://doi.org/10.1016/j.gloei.2019.07.003

Ça semble expérimenter toujours

Je continue avec l’idée que j’avais baptisée « Projet Aqueduc ». Je suis en train de préparer un article sur ce sujet, du type « démonstration de faisabilité ». Je le prépare en anglais et je me suis dit que c’est une bonne idée de reformuler en français ce que j’ai écrit jusqu’à maintenant, l’histoire de changer l’angle intellectuel, me dégourdir un peu et prendre de la distance.

Une démonstration de faisabilité suit une logique similaire à tout autre article scientifique, sauf qu’au lieu d’explorer et vérifier une hypothèse théorique du type « les choses marchent de façon ABCD, sous conditions RTYU », j’explore et vérifie l’hypothèse qu’un concept pratique, comme celui du « Projet Aqueduc », a des fondements scientifiques suffisamment solides pour que ça vaille la peine de travailler dessus et de le tester en vie réelle. Les fondements scientifiques viennent en deux couches, en quelque sorte. La couche de base consiste à passer en revue la littérature du sujet pour voir si quelqu’un a déjà décrit des solutions similaires et là, le truc c’est explorer des différentes perspectives de similarité. Similaire ne veut pas dire identique, n’est-ce pas ? Cette revue de littérature doit apporter une structure logique – un modèle – applicable à la recherche empirique, avec des variables et des paramètres constants. C’est alors que vient la couche supérieure de démonstration de faisabilité, qui consiste à conduire de la recherche empirique proprement dite avec ce modèle.    

Moi, pour le moment, j’en suis à la couche de base. Je passe donc en revue la littérature pertinente aux solutions hydrologiques et hydroélectriques, tout en formant, progressivement, un modèle numérique du « Projet Aqueduc ». Dans cette mise à jour, je commence par une brève récapitulation du concept et j’enchaîne avec ce que j’ai réussi à trouver dans la littérature. Le concept de base du « Projet Aqueduc » consiste donc à placer dans le cours d’une rivière des pompes qui travaillent selon le principe du bélier hydraulique et qui donc utilisent l’énergie cinétique de l’eau pour pomper une partie de cette eau en dehors du lit de la rivière, vers des structures marécageuses qui ont pour fonction de retenir l’eau dans l’écosystème local. Le bélier hydraulique à la capacité de pomper à la verticale aussi bien qu’à l’horizontale et donc avant d’être retenue dans les marécages, l’eau passe par une structure similaire à un aqueduc élevé (d’où le nom du concept en français), avec des réservoirs d’égalisation de flux, et ensuite elle descend vers les marécages à travers des turbines hydroélectriques. Ces dernières produisent de l’énergie qui est ensuite emmagasinée dans une installation de stockage et de là, elle est vendue pour assurer la survie financière à la structure entière. On peut ajouter des installations éoliennes et/ou photovoltaïques pour optimiser la production de l’énergie sur le terrain occupé par la structure entière.  Vous pouvez trouver une description plus élaborée du concept dans ma mise à jour intitulée « Le Catch 22 dans ce jardin d’Eden ». La faisabilité dont je veux faire une démonstration c’est la capacité de cette structure à se financer entièrement sur la base des ventes d’électricité, comme un business régulier, donc de se développer et durer sans subventions publiques. La solution pratique que je prends en compte très sérieusement en termes de créneau de vente d’électricité est une station de chargement des véhicules électriques.   

L’approche de base que j’utilise dans la démonstration de faisabilité – donc mon modèle de base – consiste à représenter le concept en question comme une chaîne des technologies :

>> TCES – stockage d’énergie

>> TCCS – station de chargement des véhicules électriques

>> TCRP – pompage en bélier hydraulique

>> TCEW – réservoirs élevés d’égalisation

>> TCCW – acheminement et siphonage d’eau

>> TCWS – l’équipement artificiel des structures marécageuses

>> TCHE – les turbines hydroélectriques

>> TCSW – installations éoliennes et photovoltaïques     

Mon intuition de départ, que j’ai l’intention de vérifier dans ma recherche à travers la littérature, est que certaines de ces technologies sont plutôt prévisibles et bien calibrées, pendant qu’il y en a d’autres qui sont plus floues et sujettes au changement, donc moins prévisibles. Les technologies prévisibles sont une sorte d’ancrage pour the concept entier et celles plus floues sont l’objet d’expérimentation.

Je commence la revue de littérature par le contexte environnemental, donc avec l’hydrologie. Les variations au niveau de la nappe phréatiques, qui est un terme scientifique pour les eaux souterraines, semblent être le facteur numéro 1 des anomalies au niveau de rétention d’eau dans les réservoirs artificiels (Neves, Nunes, & Monteiro 2020[1]). D’autre part, même sans modélisation hydrologique détaillée, il y a des preuves empiriques substantielles que la taille des réservoirs naturels et artificiels dans les plaines fluviales, ainsi que la densité de placement de ces réservoirs et ma manière de les exploiter ont une influence majeure sur l’accès pratique à l’eau dans les écosystèmes locaux. Il semble que la taille et la densité des espaces boisés intervient comme un facteur d’égalisation dans l’influence environnementale des réservoirs (Chisola, Van der Laan, & Bristow 2020[2]). Par comparaison aux autres types de technologie, l’hydrologie semble être un peu en arrière en termes de rythme d’innovation et il semble aussi que des méthodes de gestion d’innovation appliquées ailleurs avec succès peuvent marcher pour l’hydrologie, par exemple des réseaux d’innovation ou des incubateurs des technologies (Wehn & Montalvo 2018[3]; Mvulirwenande & Wehn 2020[4]). L’hydrologie rurale et agriculturale semble être plus innovatrice que l’hydrologie urbaine, par ailleurs (Wong, Rogers & Brown 2020[5]).

Ce que je trouve assez surprenant est le manque apparent de consensus scientifique à propos de la quantité d’eau dont les sociétés humaines ont besoin. Toute évaluation à ce sujet commence avec « beaucoup et certainement trop » et à partir de là, le beaucoup et le trop deviennent plutôt flous. J’ai trouvé un seul calcul, pour le moment, chez Hogeboom (2020[6]), qui maintient que la personne moyenne dans les pays développés consomme 3800 litres d’eau par jour au total, mais c’est une estimation très holistique qui inclue la consommation indirecte à travers les biens et les services ainsi que le transport. Ce qui est consommé directement via le robinet et la chasse d’eau dans les toilettes, ça reste un mystère pour la science, apparemment, à moins que la science ne considère ce sujet comment trop terre-à-terre pour s’en occuper sérieusement.     

Il y a un créneau de recherche intéressant, que certains de ses représentants appellent « la socio-hydrologie », qui étudie les comportements collectifs vis-à-vis de l’eau et des systèmes hydrologiques et qui est basée sur l’observation empirique que lesdits comportements collectifs s’adaptent, d’une façon profonde et pernicieuse à la fois, aux conditions hydrologiques que la société en question vit avec (Kumar et al. 2020[7]). Il semble que nous nous adaptons collectivement à la consommation accrue de l’eau par une productivité croissante dans l’exploitation de nos ressources hydrologiques et le revenu moyen par tête d’habitant semble être positivement corrélé avec cette productivité (Bagstad et al. 2020[8]). Il paraît donc que l’accumulation et superposition de nombreuses technologies, caractéristique aux pays développés, contribue à utiliser l’eau de façon de plus en plus productive. Dans ce contexte, il y a une recherche intéressante conduite par Mohamed et al. (2020[9]) qui avance la thèse qu’un environnement aride est non seulement un état hydrologique mais aussi une façon de gérer les ressources hydrologiques, sur ma base des données qui sont toujours incomplètes par rapport à une situation qui change rapidement.

Il y a une question qui vient plus ou moins naturellement : dans la foulée de l’adaptation socio-hydrologique quelqu’un a-t-il présenté un concept similaire à ce que moi je présente comme « Projet Aqueduc » ? Eh bien, je n’ai rien trouvé d’identique, néanmoins il y a des idées intéressement proches. Dans l’hydrologie descriptive il y a ce concept de pseudo-réservoir, qui veut dire une structure comme les marécages ou des nappes phréatiques peu profondes qui ne retiennent pas l’eau de façons statique, comme un lac artificiel, mais qui ralentissent la circulation de l’eau dans le bassin fluvial d’une rivière suffisamment pour modifier les conditions hydrologiques dans l’écosystème (Harvey et al. 2009[10]; Phiri et al. 2021[11]). D’autre part, il y a une équipe des chercheurs australiens qui ont inventé une structure qu’ils appellent par l’acronyme STORES et dont le nom complet est « short-term off-river energy storage » (Lu et al. 2021[12]; Stocks et al. 2021[13]). STORES est une structure semi-artificielle d’accumulation par pompage, où on bâtit un réservoir artificiel au sommet d’un monticule naturel placé à une certaine distance de la rivière la plus proche et ce réservoir reçoit l’eau pompée artificiellement de la rivière. Ces chercheurs australiens avancent et donnent des preuves scientifiques pour appuyer la thèse qu’avec un peu d’astuce on peut faire fonctionner ce réservoir naturel en boucle fermée avec la rivière qui l’alimente et donc de créer un système de rétention d’eau. STORES semble être relativement le plus près de mon concept de « Projet Aqueduc » et ce qui est épatant est que moi, j’avais inventé mon idée pour l’environnement des plaines alluviales de l’Europe tandis que STORES avait été mis au point pour l’environnement aride et quasi-désertique d’Australie. Enfin, il y a l’idée des soi-disant « jardins de pluie » qui sont une technologie de rétention d’eau de pluie dans l’environnement urbain, dans des structures horticulturales, souvent placées sur les toits d’immeubles (Bortolini & Zanin 2019[14], par exemple).

Je peux conclure provisoirement que tout ce qui touche à l’hydrologie strictement dite dans le cadre du « Projet Aqueduc » est sujet aux changements plutôt imprévisible. Ce que j’ai pu déduire de la littérature ressemble à un potage bouillant sous couvercle. Il y a du potentiel pour changement technologique, il y a de la pression environnementale et sociale, mais il n’y pas encore de mécanismes institutionnels récurrents pour connecter l’un à l’autre. Les technologies TCEW (réservoirs élevés d’égalisation), TCCW (acheminement et siphonage d’eau), et TCWS (l’équipement artificiel des structures marécageuses) démontrant donc un avenir flou, je passe à la technologie TCRP de pompage en bélier hydraulique. J’ai trouvé deux articles chinois, qui se suivent chronologiquement et qui semblent par ailleurs avoir été écrits par la même équipe de chercheurs : Guo et al. (2018[15]), and Li et al. (2021[16]). Ils montrent la technologie du bélier hydraulique sous un angle intéressant. D’une part, les Chinois semblent avoir donné du vrai élan à l’innovation dans ce domaine spécifique, tout au moins beaucoup plus d’élan que j’ai pu observer en Europe. D’autre part, les estimations de la hauteur effective à laquelle l’eau peut être pompée avec les béliers hydrauliques dernier cri sont respectivement de 50 mètres dans l’article de 2018 et 30 mètres dans celui de 2021. Vu que les deux articles semblent être le fruit du même projet, il y a eu comme une fascination suivie par une correction vers le bas. Quoi qu’il en soit, même l’estimation plus conservative de 30 mètres c’est nettement mieux que les 20 mètres que j’assumais jusqu’à maintenant.

Cette élévation relative possible à atteindre avec la technologie du bélier hydraulique est importante pour la technologie suivante de ma chaîne, donc celle des petites turbines hydroélectriques, la TCHE. L’élévation relative de l’eau et le flux par seconde sont les deux paramètres clés qui déterminent la puissance électrique produite (Cai, Ye & Gholinia 2020[17]) et il se trouve que dans le « Projet Aqueduc », avec l’élévation et le flux largement contrôlés à travers la technologie du bélier hydraulique, les turbines deviennent un peu moins dépendantes sur les conditions naturelles.

J’ai trouvé une revue merveilleusement encyclopédique des paramètres pertinents aux petites turbines hydroélectriques chez Hatata, El-Saadawi, & Saad (2019[18]). La puissance électrique se calcule donc comme : Puissance = densité de l’eau (1000 kg/m3) * constante d’accélération gravitationnelle (9,8 m/s2) * élévation nette (mètres) * Q (flux par seconde m3/s).

L’investissement initial en de telles installations se calcule par unité de puissance, donc sur la base de 1 kilowatt et se divise en 6 catégories : la construction de la prise d’eau, la centrale électrique strictement dite, les turbines, le générateur, l’équipement auxiliaire, le transformateur et enfin le poste extérieur. Je me dis par ailleurs que – vu la structure du « Projet Aqueduc » – l’investissement en la construction de prise d’eau est en quelque sorte équivalent au système des béliers hydrauliques et réservoirs élevés. En tout cas :

>> la construction de la prise d’eau, par 1 kW de puissance  ($) 186,216 * Puissance-0,2368 * Élévation -0,597

>> la centrale électrique strictement dite, par 1 kW de puissance  ($) 1389,16 * Puissance-0,2351 * Élévation-0,0585

>> les turbines, par 1 kW de puissance  ($)

@ la turbine Kaplan: 39398 * Puissance-0,58338 * Élévation-0,113901

@ la turbine Frances: 30462 * Puissance-0,560135 * Élévation-0,127243

@ la turbine à impulsions radiales: 10486,65 * Puissance-0,3644725 * Élévation-0,281735

@ la turbine Pelton: 2 * la turbine à impulsions radiales

>> le générateur, par 1 kW de puissance  ($) 1179,86 * Puissance-0,1855 * Élévation-0,2083

>> l’équipement auxiliaire, par 1 kW de puissance  ($) 612,87 * Puissance-0,1892 * Élévation-0,2118

>> le transformateur et le poste extérieur, par 1 kW de puissance 

($) 281 * Puissance0,1803 * Élévation-0,2075

Une fois la puissance électrique calculée avec le paramètre d’élévation relative assurée par les béliers hydrauliques, je peux calculer l’investissement initial en hydro-génération comme la somme des positions mentionnées ci-dessus. Hatata, El-Saadawi, & Saad (2019 op. cit.) recommandent aussi de multiplier une telle somme par le facteur de 1,13 (c’est donc un facteur du type « on ne sait jamais ») et d’assumer que les frais courants d’exploitation annuelle vont se situer entre 1% et 6% de l’investissement initial.

Syahputra & Soesanti (2021[19]) étudient le cas de la rivière Progo, dotée d’un flux tout à fait modeste de 6,696 mètres cubes par seconde et située dans Kulon Progo Regency (une region spéciale au sein de Yogyakarta, Indonesia). Le système des petites turbines hydroélectriques y fournit l’électricité aux 962 ménages locaux, et crée un surplus de 4 263 951 kWh par an d’énergie à revendre aux consommateurs externes. Dans un autre article, Sterl et al. (2020[20]) étudient le cas de Suriname et avancent une thèse intéressante, notamment que le développement d’installations basées sur les énergies renouvelables crée un phénomène d’appétit d’énergie qui croît à mesure de manger et qu’un tel développement en une source d’énergie – le vent, par exemple – stimule l’investissement en installations basées sur d’autres sources, donc l’hydraulique et le photovoltaïque.  

Ces études relativement récentes corroborent celles d’il y a quelques années, comme celle de Vilanova & Balestieri (2014[21]) ou bien celle de Vieira et al. (2015[22]), avec une conclusion générale que les petites turbines hydroélectriques ont atteint un degré de sophistication technologique suffisante pour dégager une quantité d’énergie économiquement profitable. Par ailleurs, il semble qu’il y a beaucoup à gagner dans ce domaine à travers l’optimisation de la distribution de puissance entre les turbines différentes. De retour aux publications les plus récentes, j’ai trouvé des études de faisabilité tout à fait robustes pour les petites turbines hydroélectriques, qui indiquent que – pourvu qu’on soit prêt à accepter un retour d’environ 10 à 11 ans sur l’investissement initial – le petit hydro peut être exploité profitablement même avec une élévation relative en dessous de 20 mètres (Arthur et al. 2020[23] ; Ali et al. 2021[24]).

C’est ainsi que j’arrive donc à la portion finale dans la chaîne technologique du « Projet Aqueduc », donc au stockage d’énergie (TCES) ainsi que TCCS ou la station de chargement des véhicules électriques. La puissance à installer dans une station de chargement semble se situer entre 700 et 1000 kilowatts (Zhang et al. 2018[25]; McKinsey 2018[26]). En dessous de 700 kilowatt la station peut devenir si difficile à accéder pour le consommateur moyen, due aux files d’attente, qu’elle peut perdre la confiance des clients locaux. En revanche, tout ce qui va au-dessus de 1000 kilowatts est vraiment utile seulement aux heures de pointe dans des environnements urbains denses. Il y a des études de concept pour les stations de chargement où l’unité de stockage d’énergie est alimentée à partir des sources renouvelables (Al Wahedi & Bicer 2020[27]). Zhang et al. (2019[28]) présentent un concept d’entreprise tout fait pour une station de chargement située dans le milieu urbain. Apparemment, le seuil de profitabilité se situe aux environs de 5 100 000 kilowatt heures vendues par an.  

En termes de technologie de stockage strictement dite, les batteries Li-ion semblent être la solution de base pour maintenant, quoi qu’une combinaison avec les piles à combustible ou bien avec l’hydrogène semble prometteuse (Al Wahedi & Bicer 2020 op. cit. ; Sharma, Panvar & Tripati 2020[29]). En général, pour le moment, les batteries Li-Ion montrent le rythme d’innovation relativement le plus soutenu (Tomaszewska et al. 2019[30] ; de Simone & Piegari 2019[31]; Koohi-Fayegh & Rosen 2020[32]). Un article récent par Elmeligy et al. (2021[33]) présente un concept intéressant d’unité mobile de stockage qui pourrait se déplacer entre plusieurs stations de chargement. Quant à l’investissement initial requis pour une station de chargement, ça semble expérimenter toujours mais la marge de manœuvre se rétrécit pour tomber quelque part entre $600 ÷ $800 par 1 kW de puissance (Cole & Frazier 2019[34]; Cole, Frazier, Augustine 2021[35]).


[1] Neves, M. C., Nunes, L. M., & Monteiro, J. P. (2020). Evaluation of GRACE data for water resource management in Iberia: a case study of groundwater storage monitoring in the Algarve region. Journal of Hydrology: Regional Studies, 32, 100734. https://doi.org/10.1016/j.ejrh.2020.100734

[2] Chisola, M. N., Van der Laan, M., & Bristow, K. L. (2020). A landscape hydrology approach to inform sustainable water resource management under a changing environment. A case study for the Kaleya River Catchment, Zambia. Journal of Hydrology: Regional Studies, 32, 100762. https://doi.org/10.1016/j.ejrh.2020.100762

[3] Wehn, U., & Montalvo, C. (2018). Exploring the dynamics of water innovation: Foundations for water innovation studies. Journal of Cleaner Production, 171, S1-S19. https://doi.org/10.1016/j.jclepro.2017.10.118

[4] Mvulirwenande, S., & Wehn, U. (2020). Fostering water innovation in Africa through virtual incubation: Insights from the Dutch VIA Water programme. Environmental Science & Policy, 114, 119-127. https://doi.org/10.1016/j.envsci.2020.07.025

[5] Wong, T. H., Rogers, B. C., & Brown, R. R. (2020). Transforming cities through water-sensitive principles and practices. One Earth, 3(4), 436-447. https://doi.org/10.1016/j.oneear.2020.09.012

[6] Hogeboom, R. J. (2020). The Water Footprint Concept and Water’s Grand Environmental Challenges. One earth, 2(3), 218-222. https://doi.org/10.1016/j.oneear.2020.02.010

[7] Kumar, P., Avtar, R., Dasgupta, R., Johnson, B. A., Mukherjee, A., Ahsan, M. N., … & Mishra, B. K. (2020). Socio-hydrology: A key approach for adaptation to water scarcity and achieving human well-being in large riverine islands. Progress in Disaster Science, 8, 100134. https://doi.org/10.1016/j.pdisas.2020.100134

[8] Bagstad, K. J., Ancona, Z. H., Hass, J., Glynn, P. D., Wentland, S., Vardon, M., & Fay, J. (2020). Integrating physical and economic data into experimental water accounts for the United States: Lessons and opportunities. Ecosystem Services, 45, 101182. https://doi.org/10.1016/j.ecoser.2020.101182

[9] Mohamed, M. M., El-Shorbagy, W., Kizhisseri, M. I., Chowdhury, R., & McDonald, A. (2020). Evaluation of policy scenarios for water resources planning and management in an arid region. Journal of Hydrology: Regional Studies, 32, 100758. https://doi.org/10.1016/j.ejrh.2020.100758

[10] Harvey, J.W., Schaffranek, R.W., Noe, G.B., Larsen, L.G., Nowacki, D.J., O’Connor, B.L., 2009. Hydroecological factors governing surface water flow on a low-gradient floodplain. Water Resour. Res. 45, W03421, https://doi.org/10.1029/2008WR007129.

[11] Phiri, W. K., Vanzo, D., Banda, K., Nyirenda, E., & Nyambe, I. A. (2021). A pseudo-reservoir concept in SWAT model for the simulation of an alluvial floodplain in a complex tropical river system. Journal of Hydrology: Regional Studies, 33, 100770. https://doi.org/10.1016/j.ejrh.2020.100770.

[12] Lu, B., Blakers, A., Stocks, M., & Do, T. N. (2021). Low-cost, low-emission 100% renewable electricity in Southeast Asia supported by pumped hydro storage. Energy, 121387. https://doi.org/10.1016/j.energy.2021.121387

[13] Stocks, M., Stocks, R., Lu, B., Cheng, C., & Blakers, A. (2021). Global atlas of closed-loop pumped hydro energy storage. Joule, 5(1), 270-284. https://doi.org/10.1016/j.joule.2020.11.015

[14] Bortolini, L., & Zanin, G. (2019). Reprint of: Hydrological behaviour of rain gardens and plant suitability: A study in the Veneto plain (north-eastern Italy) conditions. Urban forestry & urban greening, 37, 74-86. https://doi.org/10.1016/j.ufug.2018.07.003

[15] Guo, X., Li, J., Yang, K., Fu, H., Wang, T., Guo, Y., … & Huang, W. (2018). Optimal design and performance analysis of hydraulic ram pump system. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Puissance and Energy, 232(7), 841-855. https://doi.org/10.1177%2F0957650918756761

[16] Li, J., Yang, K., Guo, X., Huang, W., Wang, T., Guo, Y., & Fu, H. (2021). Structural design and parameter optimization on a waste valve for hydraulic ram pumps. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Puissance and Energy, 235(4), 747–765. https://doi.org/10.1177/0957650920967489

[17] Cai, X., Ye, F., & Gholinia, F. (2020). Application of artificial neural network and Soil and Water Assessment Tools in evaluating Puissance generation of small hydroPuissance stations. Energy Reports, 6, 2106-2118. https://doi.org/10.1016/j.egyr.2020.08.010.

[18] Hatata, A. Y., El-Saadawi, M. M., & Saad, S. (2019). A feasibility study of small hydro Puissance for selected locations in Egypt. Energy Strategy Reviews, 24, 300-313. https://doi.org/10.1016/j.esr.2019.04.013

[19] Syahputra, R., & Soesanti, I. (2021). Renewable energy systems based on micro-hydro and solar photovoltaic for rural areas: A case study in Yogyakarta, Indonesia. Energy Reports, 7, 472-490. https://doi.org/10.1016/j.egyr.2021.01.015

[20] Sterl, S., Donk, P., Willems, P., & Thiery, W. (2020). Turbines of the Caribbean: Decarbonising Suriname’s electricity mix through hydro-supported integration of wind Puissance. Renewable and Sustainable Energy Reviews, 134, 110352. https://doi.org/10.1016/j.rser.2020.110352

[21] Vilanova, M. R. N., & Balestieri, J. A. P. (2014). HydroPuissance recovery in water supply systems: Models and case study. Energy conversion and management, 84, 414-426. https://doi.org/10.1016/j.enconman.2014.04.057

[22] Vieira, D. A. G., Guedes, L. S. M., Lisboa, A. C., & Saldanha, R. R. (2015). Formulations for hydroelectric energy production with optimality conditions. Energy Conversion and Management, 89, 781-788. https://doi.org/10.1016/j.enconman.2014.10.048

[23] Arthur, E., Anyemedu, F. O. K., Gyamfi, C., Asantewaa-Tannor, P., Adjei, K. A., Anornu, G. K., & Odai, S. N. (2020). Potential for small hydroPuissance development in the Lower Pra River Basin, Ghana. Journal of Hydrology: Regional Studies, 32, 100757. https://doi.org/10.1016/j.ejrh.2020.100757

[24] Ali, M., Wazir, R., Imran, K., Ullah, K., Janjua, A. K., Ulasyar, A., … & Guerrero, J. M. (2021). Techno-economic assessment and sustainability impact of hybrid energy systems in Gilgit-Baltistan, Pakistan. Energy Reports, 7, 2546-2562. https://doi.org/10.1016/j.egyr.2021.04.036

[25] Zhang, Y., He, Y., Wang, X., Wang, Y., Fang, C., Xue, H., & Fang, C. (2018). Modeling of fast charging station equipped with energy storage. Global Energy Interconnection, 1(2), 145-152. DOI:10.14171/j.2096-5117.gei.2018.02.006

[26] McKinsey Center for Future Mobility, How Battery Storage Can Help Charge the Electric-Vehicle Market?, February 2018,

[27] Al Wahedi, A., & Bicer, Y. (2020). Development of an off-grid electrical vehicle charging station hybridized with renewables including battery cooling system and multiple energy storage units. Energy Reports, 6, 2006-2021. https://doi.org/10.1016/j.egyr.2020.07.022

[28] Zhang, J., Liu, C., Yuan, R., Li, T., Li, K., Li, B., … & Jiang, Z. (2019). Design scheme for fast charging station for electric vehicles with distributed photovoltaic power generation. Global Energy Interconnection, 2(2), 150-159. https://doi.org/10.1016/j.gloei.2019.07.003

[29] Sharma, S., Panwar, A. K., & Tripathi, M. M. (2020). Storage technologies for electric vehicles. Journal of traffic and transportation engineering (english edition), 7(3), 340-361. https://doi.org/10.1016/j.jtte.2020.04.004

[30] Tomaszewska, A., Chu, Z., Feng, X., O’Kane, S., Liu, X., Chen, J., … & Wu, B. (2019). Lithium-ion battery fast charging: A review. ETransportation, 1, 100011. https://doi.org/10.1016/j.etran.2019.100011

[31] De Simone, D., & Piegari, L. (2019). Integration of stationary batteries for fast charge EV charging stations. Energies, 12(24), 4638. https://doi.org/10.3390/en12244638

[32] Koohi-Fayegh, S., & Rosen, M. A. (2020). A review of energy storage types, applications and recent developments. Journal of Energy Storage, 27, 101047. https://doi.org/10.1016/j.est.2019.101047

[33] Elmeligy, M. M., Shaaban, M. F., Azab, A., Azzouz, M. A., & Mokhtar, M. (2021). A Mobile Energy Storage Unit Serving Multiple EV Charging Stations. Energies, 14(10), 2969. https://doi.org/10.3390/en14102969

[34] Cole, Wesley, and A. Will Frazier. 2019. Cost Projections for Utility-Scale Battery Storage.

Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-73222. https://www.nrel.gov/docs/fy19osti/73222.pdf

[35] Cole, Wesley, A. Will Frazier, and Chad Augustine. 2021. Cost Projections for UtilityScale Battery Storage: 2021 Update. Golden, CO: National Renewable Energy

Laboratory. NREL/TP-6A20-79236. https://www.nrel.gov/docs/fy21osti/79236.pdf.

We keep going until we observe

I keep working on a proof-of-concept paper for my idea of ‘Energy Ponds’. In my last two updates, namely in ‘Seasonal lakes’, and in ‘Le Catch 22 dans ce jardin d’Eden’, I sort of refreshed my ideas and set the canvas for painting. Now, I start sketching. What exact concept do I want to prove, and what kind of evidence can possibly confirm (or discard) that concept? The idea I am working on has a few different layers. The most general vision is that of purposefully storing water in spongy structures akin to swamps or wetlands. These can bear various degree of artificial construction, and can stretch from natural wetlands, through semi-artificial ones, all the way to urban technologies such as rain gardens and sponge cities. The most general proof corresponding to that vision is a review of publicly available research – peer-reviewed papers, preprints, databases etc. – on that general topic.

Against that general landscape, I sketch two more specific concepts: the idea of using ram pumps as a technology of forced water retention, and the possibility of locating those wetland structures in the broadly spoken Northern Europe, thus my home region. Correspondingly, I need to provide two streams of scientific proof: a review of literature on the technology of ram pumping, on the one hand, and on the actual natural conditions, as well as land management policies in Europe, on the other hand.  I need to consider the environmental impact of creating new wetland-like structures in Northern Europe, as well as the socio-economic impact, and legal feasibility of conducting such projects.

Next, I sort of build upwards. I hypothesise a complex technology, where ram-pumped water from the river goes into a sort of light elevated tanks, and from there, using the principle of Roman siphon, cascades down into wetlands, and through a series of small hydro-electric turbines. Turbines generate electricity, which is being stored and then sold outside.

At that point, I have a technology of water retention coupled with a technology of energy generation and storage. I further advance a second hypothesis that such a complex technology will be economically sustainable based on the corresponding sales of electricity. In other words, I want to figure out a configuration of that technology, which will be suitable for communities which either don’t care at all, or simply cannot afford to care about the positive environmental impact of the solution proposed.

Proof of concept for those two hypotheses is going to be complex. First, I need to pass in review the available technologies for energy storage, energy generation, as well as for the construction of elevated tanks and Roman siphons. I need to take into account various technological mixes, including the incorporation of wind turbines and photovoltaic installation into the whole thing, in order to optimize the output of energy. I will try to look for documented examples of small hydro-generation coupled with wind and solar. Then, I have to rack the literature as regards mathematical models for the optimization of such power systems and put them against my own idea of reverse engineering back from the storage technology. I take the technology of energy storage which seems the most suitable for the local market of energy, and for the hypothetical charging from hydro-wind-solar mixed generation. I build a control scenario where that storage facility just buys energy at wholesale prices from the power grid and then resells it. Next, I configure the hydro-wind-solar generation so as to make it economically competitive against the supply of energy from the power grid.

Now, I sketch. I keep in mind the levels of conceptualization outlined above, and I quickly move through published science along that logical path, quickly picking a few articles for each topic. I am going to put those nonchalantly collected pieces of science back-to-back and see how and whether at all it all makes sense together. I start with Bortolini & Zanin (2019[1]), who study the impact of rain gardens on water management in cities of the Veneto region in Italy. Rain gardens are vegetal structures, set up in the urban environment, with the specific purpose to retain rainwater.  Bortolini & Zanin (2019 op. cit.) use a simplified water balance, where the rain garden absorbs and retains a volume ‘I’ of water (‘I’ stands for infiltration), which is the difference between precipitations on the one hand, and the sum total of overflowing runoff from the rain garden plus evapotranspiration of water, on the other hand. Soil and plants in the rain garden have a given top capacity to retain water. Green plants typically hold 80 – 95% of their mass in water, whilst trees hold about 50%. Soil is considered wet when it contains about 25% of water. The rain garden absorbs water from precipitations at a rate determined by hydraulic conductivity, which means the relative ease of a fluid (usually water) to move through pore spaces or fractures, and which depends on the intrinsic permeability of the material, the degree of saturation, and on the density and viscosity of the fluid.

As I look at it, I can see that the actual capacity of water retention in a rain garden can hardly be determined a priori, unless we have really a lot of empirical data from the given location. For a new location of a new rain garden, it is safe to assume that we need an experimental phase when we empirically assess the retentive capacity of the rain garden with different configurations of soil and vegetation used. That leads me to generalizing that any porous structure we use for retaining rainwater, would it be something like wetlands, or something like a rain garden in urban environment, has a natural constraint of hydraulic conductivity, and that constraint determines the percentage of precipitations, and the metric volume thereof, which the given structure can retain.

Bortolini & Zanin (2019 op. cit.) bring forth empirical results which suggest that properly designed rain gardens located on rooftops in a city can absorb from 87% to 93% of the total input of water they receive. Cool. I move on and towards the issue of water management in Europe, with a working paper by Fribourg-Blanc, B. (2018[2]), and the most important takeaway from that paper is that we have something called European Platform for Natural Water Retention Measures AKA http://nwrm.eu , and that thing have both good properties and bad properties. The good thing about http://nwrm.eu is that it contains loads of data and publications about projects in Natural Water Retention in Europe. The bad thing is that http://nwrm.eu is not a secure website. Another paper, by Tóth et al. (2017[3]) tells me that another analytical tool exists, namely the European Soil Hydraulic Database (EU‐ SoilHydroGrids ver1.0).

So far, so good. I already know there is data and science for evaluating, with acceptable precision, the optimal structure and the capacity for water retention in porous structures such as rain gardens or wetlands, in the European context. I move to the technology of ram pumps. I grab two papers: Guo et al. (2018[4]), and Li et al. (2021[5]). They show me two important things. Firstly, China seems to be burning the rubber in the field of ram pumping technology. Secondly, the greatest uncertainty as for that technology seems to be the actual height those ram pumps can elevate water at, or, when coupled with hydropower, the hydraulic head which ram pumps can create. Guo et al. (2018 op. cit.) claim that 50 meters of elevation is the maximum which is both feasible and efficient. Li et al. (2021 op. cit.) are sort of vertically more conservative and claim that the whole thing should be kept below 30 meters of elevation. Both are better than 20 meters, which is what I thought was the best one can expect. Greater elevation of water means greater hydraulic head, and more hydropower to be generated. It pays off to review literature.

Lots of uncertainty as for the actual capacity and efficiency of ram pumping means quick technological change in that domain. This is economically interesting. It means that investing in projects which involve ram pumping means investment in quickly changing a technology. That means both high hopes for an even better technology in immediate future, and high needs for cash in the balance sheet of the entities involved.

I move to the end-of-the-pipeline technology in my concept, namely to energy storage. I study a paper by Koohi-Fayegh & Rosen (2020[6]), which suggests two things. Firstly, for a standalone installation in renewable energy, whatever combination of small hydropower, photovoltaic and small wind turbines we think of, lithium-ion batteries are always a good idea for power storage, Secondly, when we work with hydrogeneration, thus when we have any hydraulic head to make electricity with, pumped storage comes sort of natural. That leads me to an idea which looks even crazier than what I have imagined so far: what if we create an elevated garden with strong capacity for water retention. Ram pumps take water from the river and pump it up onto elevated platforms with rain gardens on it. Those platforms can be optimized as for their absorption of sunlight and thus as regards their interaction with whatever is underneath them.  

I move to small hydro, and I find two papers, namely Couto & Olden (2018[7]), and Lange et al. (2018[8]), which are both interestingly critical as regards small hydropower installations. Lange et al. (2018 op. cit.) claim that the overall environmental impact of small hydro should be closely monitored. Couto & Olden (2018 op. cit.) go further and claim there is a ‘craze’ about small hydro, and that craze has already lead to overinvestment in the corresponding installations, which can be damaging both environmentally and economically (overinvestment means financial collapse of many projects). Those critical views in mind, I turn to another paper, by Zhou et al. (2019[9]), who approach the issue as a case for optimization, within a broader framework called ‘Water-Food-Energy’ Nexus, WFE for closer friends. This paper, just as a few others it cites (Ming et al. 2018[10]; Uen et al. 2018[11]), advocates for using artificial intelligence in order to optimize for WFE.

Zhou et al. (2019 op.cit.) set three hydrological scenarios for empirical research and simulation. The baseline scenario corresponds to an average hydrological year, with average water levels and average precipitations. Next to it are: a dry year and a wet year. The authors assume that the cost of installation in small hydropower is $600 per kW on average.  They simulate the use of two technologies for hydro-electric turbines: Pelton and Vortex. Pelton turbines are optimized paddled wheels, essentially, whilst the Vortex technology consists in creating, precisely, a vortex of water, and that vortex moves a rotor placed in the middle of it.

Zhou et al. (2019 op.cit.) create a multi-objective function to optimize, with the following desired outcomes:

>> Objective 1: maximize the reliability of water supply by minimizing the probability of real water shortage occurring.

>> Objective 2: maximize water storage given the capacity of the reservoir. Note: reservoir is understood hydrologically, as any structure, natural or artificial, able to retain water.

>> Objective 3: maximize the average annual output of small hydro-electric turbines

Those objectives are being achieved under the corresponding sets of constraints. For water supply those constraints all turn around water balance, whilst for energy output it is more about the engineering properties of the technologies taken into account. The three objectives are hierarchized. First, Zhou et al. (2019 op.cit.) perform an optimization regarding Objectives 1 and 2, thus in order to find the optimal hydrological characteristics to meet, and then, on the basis of these, they optimize the technology to put in place, as regards power output.

The general tool for optimization used by Zhou et al. (2019 op.cit.) is a genetic algorithm called NSGA-II, AKA Non-dominated Sorting Genetic Algorithm. Apparently, NSGA-II has a long and successful history of good track in engineering, including water management and energy (see e.g. Chang et al. 2016[12]; Jain & Sachdeva 2017[13];  Assaf & Shabani 2018[14]). I want to stop for a while here and have a good look at this specific algorithm. The logic of NSGA-II starts with creating an initial population of cases/situations/configurations etc. Each case is a combination of observations as regards the objectives to meet, and the actual values observed in constraining variables, e.g. precipitations for water balance or hydraulic head for the output of hydropower. In the conventional lingo of this algorithm, those cases are called chromosomes. Yes, I know, a hydro-electric turbine placed in the context of water management hardly looks like a chromosome, but it is a genetic algorithm, and it just sounds fancy to use that biologically marked vocabulary.

As for me, I like staying close to real life, and therefore I call those cases solutions rather than chromosomes. Anyway, the underlying math is the same. Once I have that initial population of real-life solutions, I calculate two parameters for each of them: their rank as regards the objectives to maximize, and their so-called ‘crowded distance’. Ranking is done with the procedure of fast non-dominated sorting. It is a comparison in pairs, where the solution A dominates another solution B, if and only if there is no objective of A worse than that objective of B and there is at least one objective of A better than that objective of B. The solution which scores the most wins in such peer-to-peer comparisons is at the top of the ranking, the one with the second score of wins is the second etc. Crowding distance is essentially the same as what I call coefficient of coherence in my own research: Euclidean distance (or other mathematical distance) is calculated for each pair of solutions. As a result, each solution is associated with k Euclidean distances to the k remaining solutions, which can be reduced to an average distance, i.e. the crowded distance.

In the next step, an off-spring population is produced from that original population of solutions. It is created by taking relatively the fittest solutions from the initial population, recombining their characteristics in a 50/50 proportion, and adding them some capacity for endogenous mutation. Two out of these three genetic functions are de facto controlled. We choose relatively the fittest by establishing some kind of threshold for fitness, as regards the objectives pursued. It can be a required minimum, a quantile (e.g. the third quartile), or an average. In the first case, we arbitrarily impose a scale of fitness on our population, whilst in the latter two the hierarchy of fitness is generated endogenously from the population of solutions observed. Fitness can have shades and grades, by weighing the score in non-dominated sorting, thus the number of wins over other solutions, on the one hand, and the crowded distance on the other hand. In other words, we can go for solutions which have a lot of similar ones in the population (i.e. which have a low average crowded distance), or, conversely, we can privilege lone wolves, with a high average Euclidean distance from anything else on the plate.  

The capacity for endogenous mutation means that we can allow variance in all or in just the selected variables which make each solution. The number of degrees of freedom we allow in each variable dictates the number of mutations that can be created. Once again, discreet power is given to the analyst: we can choose the genetic traits which can mutate and we can determine their freedom to mutate. In an engineering problem, technological and environmental constraints should normally put a cap on the capacity for mutation. Still, we can think about an algorithm which definitely kicks the lid off the barrel of reality, and which generates mutations in the wildest registers of variables considered. It is a way to simulate a process when the presence of strong outliers has a strong impact on the whole population.

The same discreet cap on the freedom to evolve is to be found when we repeat the process. The offspring generation of solutions goes essentially through the same process as the initial one, to produce further offspring: ranking by non-dominated sorting and crowded distance, selection of the fittest, recombination, and endogenous mutation. At the starting point of this process, we can be two alternative versions of the Mother Nature. We can be a mean Mother Nature, and we shave off from the offspring population all those baby-solutions which do not meet the initial constraints, e.g. zero supply of water in this specific case. On the other hand, we can be even meaner a Mother Nature and allow those strange, dysfunctional mutants to keep going and see what happens to the whole species after a few rounds of genetic reproduction.

With each generation, we compute an average crowded distance between all the solutions created, i.e. we check how diverse is the species in this generation. As long as diversity grows or remains constant, we assume that the divergence between the solutions generated grows or stays the same. Similarly, we can compute an even more general crowded distance between each pair of generations, and therefore to assess how far has the current generation gone from the parent one. We keep going until we observe that the intra-generational crowded distance and the inter-generational one start narrowing down asymptotically to zero. In other words, we consider resuming evolution when solutions in the game become highly similar to each other and when genetic change stops bringing significant functional change.

Cool. When I want to optimize my concept of Energy Ponds, I need to add the objective of constrained return on investment, based on the sales of electricity. In comparison to Zhou et al. (2019 op.cit.), I need to add a third level of selection. I start with selecting environmentally the solutions which make sense in terms of water management. In the next step, I produce a range of solutions which assure the greatest output of power, in a possible mix with solar and wind. Then I take those and filter them through the NSGA-II procedure as regards their capacity to sustain themselves financially. Mind you, I can shake it off a bit by fusing together those levels of selection. I can simulate extreme cases, when, for example, good economic sustainability becomes an environmental problem. Still, it would be rather theoretical. In Europe, non-compliance with environmental requirements makes a project a non-starter per se: you just can get the necessary permits if your hydropower project messes with hydrological constraints legally imposed on the given location.     

Cool. It all starts making sense. There is apparently a lot of stir in the technology of making semi-artificial structures for retaining water, such as rain gardens and wetlands. That means a lot of experimentation, and that experimentation can be guided and optimized by testing the fitness of alternative solutions for meeting objectives of water management, power output and economic sustainability. I have some starting data, to produce the initial generation of solutions, and then try to optimize them with an algorithm such as NSGA-II.


[1] Bortolini, L., & Zanin, G. (2019). Reprint of: Hydrological behaviour of rain gardens and plant suitability: A study in the Veneto plain (north-eastern Italy) conditions. Urban forestry & urban greening, 37, 74-86. https://doi.org/10.1016/j.ufug.2018.07.003

[2] Fribourg-Blanc, B. (2018, April). Natural Water Retention Measures (NWRM), a tool to manage hydrological issues in Europe?. In EGU General Assembly Conference Abstracts (p. 19043). https://ui.adsabs.harvard.edu/abs/2018EGUGA..2019043F/abstract

[3] Tóth, B., Weynants, M., Pásztor, L., & Hengl, T. (2017). 3D soil hydraulic database of Europe at 250 m resolution. Hydrological Processes, 31(14), 2662-2666. https://onlinelibrary.wiley.com/doi/pdf/10.1002/hyp.11203

[4] Guo, X., Li, J., Yang, K., Fu, H., Wang, T., Guo, Y., … & Huang, W. (2018). Optimal design and performance analysis of hydraulic ram pump system. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy, 232(7), 841-855. https://doi.org/10.1177%2F0957650918756761

[5] Li, J., Yang, K., Guo, X., Huang, W., Wang, T., Guo, Y., & Fu, H. (2021). Structural design and parameter optimization on a waste valve for hydraulic ram pumps. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy, 235(4), 747–765. https://doi.org/10.1177/0957650920967489

[6] Koohi-Fayegh, S., & Rosen, M. A. (2020). A review of energy storage types, applications and recent developments. Journal of Energy Storage, 27, 101047. https://doi.org/10.1016/j.est.2019.101047

[7] Couto, T. B., & Olden, J. D. (2018). Global proliferation of small hydropower plants–science and policy. Frontiers in Ecology and the Environment, 16(2), 91-100. https://doi.org/10.1002/fee.1746

[8] Lange, K., Meier, P., Trautwein, C., Schmid, M., Robinson, C. T., Weber, C., & Brodersen, J. (2018). Basin‐scale effects of small hydropower on biodiversity dynamics. Frontiers in Ecology and the Environment, 16(7), 397-404.  https://doi.org/10.1002/fee.1823

[9] Zhou, Y., Chang, L. C., Uen, T. S., Guo, S., Xu, C. Y., & Chang, F. J. (2019). Prospect for small-hydropower installation settled upon optimal water allocation: An action to stimulate synergies of water-food-energy nexus. Applied Energy, 238, 668-682. https://doi.org/10.1016/j.apenergy.2019.01.069

[10] Ming, B., Liu, P., Cheng, L., Zhou, Y., & Wang, X. (2018). Optimal daily generation scheduling of large hydro–photovoltaic hybrid power plants. Energy Conversion and Management, 171, 528-540. https://doi.org/10.1016/j.enconman.2018.06.001

[11] Uen, T. S., Chang, F. J., Zhou, Y., & Tsai, W. P. (2018). Exploring synergistic benefits of Water-Food-Energy Nexus through multi-objective reservoir optimization schemes. Science of the Total Environment, 633, 341-351. https://doi.org/10.1016/j.scitotenv.2018.03.172

[12] Chang, F. J., Wang, Y. C., & Tsai, W. P. (2016). Modelling intelligent water resources allocation for multi-users. Water resources management, 30(4), 1395-1413. https://doi.org/10.1007/s11269-016-1229-6

[13] Jain, V., & Sachdeva, G. (2017). Energy, exergy, economic (3E) analyses and multi-objective optimization of vapor absorption heat transformer using NSGA-II technique. Energy Conversion and Management, 148, 1096-1113. https://doi.org/10.1016/j.enconman.2017.06.055

[14] Assaf, J., & Shabani, B. (2018). Multi-objective sizing optimisation of a solar-thermal system integrated with a solar-hydrogen combined heat and power system, using genetic algorithm. Energy Conversion and Management, 164, 518-532. https://doi.org/10.1016/j.enconman.2018.03.026

Seasonal lakes

Once again, been a while since I last blogged. What do you want, I am having a busy summer. Putting order in my own chaos, and, over the top of that, putting order in other people’s chaos, this is all quite demanding in terms of time and energy. What? Without trying to put order in chaos, that chaos might take less time and energy? Well, yes, but order look tidier than chaos.

I am returning to the technological concept which I labelled ‘Energy Ponds’ (or ‘projet Aqueduc’ in French >> see: Le Catch 22 dans ce jardin d’Eden). You can find a description of that concept onder the hyperlinked titles provided. I am focusing on refining my repertoire of skills in scientific validation of technological concepts. I am passing in review some recent literature, and I am trying to find good narrative practices in that domain.

The general background of ‘Energy Ponds’ consists in natural phenomena observable in Europe as the climate change progresses, namely: a) long-term shift in the structure of precipitations, from snow to rain b) increasing occurrence of floods and droughts c) spontaneous reemergence of wetlands. All these phenomena have one common denominator: increasingly volatile flow per second in rivers. The essential idea of Energy Ponds is to ‘financialize’ that volatile flow, so to say, i.e. to capture its local surpluses, store them for later, and use the very mechanism of storage itself as a source of economic value.

When water flows downstream, in a river, its retention can be approached as the opportunity for the same water to loop many times over the same specific portion of the collecting basin (of the river). Once such a loop is created, we can extend the average time that a liter of water spends in the whereabouts. Ram pumps, connected to storage structures akin to swamps, can give such an opportunity. A ram pump uses the kinetic energy of flowing water in order to pump some of that flow up and away from its mainstream. Ram pumps allow forcing a process, which we now as otherwise natural. Rivers, especially in geological plains, where they flow relatively slowly, tend to build, with time, multiple ramifications. Those branchings can be directly observable at the surface, as meanders, floodplains or seasonal lakes, but much of them is underground, as pockets of groundwater. In this respect, it is useful to keep in mind that mechanically, rivers are the drainpipes of rainwater from their respective basins. Another basic hydrological fact, useful to remember in the context of the Energy Ponds concept, is that strictly speaking retention of rainwater – i.e. a complete halt in its circulation through the collecting basin of the river – is rarely possible, and just as rarely it is a sensible idea to implement. Retention means rather a slowdown to the flow of rainwater through the collecting basin into the river.

One of the ways that water can be slowed down consists in making it loop many times over the same section of the river. Let’s imagine a simple looping sequence: water from the river is being ram-pumped up and away into retentive structures akin to swamps, i.e. moderately deep spongy structures underground, with high capacity for retention, covered with a superficial layer of shallow-rooted vegetation. With time, as the swamp fills with water, the surplus is evacuated back into the river, by a system of canals. Water stored in the swamp will be ultimately evacuated, too, minus evaporation, it will just happen much more slowly, by the intermediary of groundwaters. In order to illustrate the concept mathematically, let’ s suppose that we have water in the river flowing at the pace of, e.g. 45 m3 per second. We make it loop once via ram pumps and retentive swamps, and, if as a result of that looping, the speed of the flow is sliced by 3. On the long run we slow down the way that the river works as the local drainpipe: we slow it from 43 m3 per second down to [43/3 = 14,33…] m3 per second.  As water from the river flows slower overall, it can yield more environmental services: each cubic meter of water has more time to ‘work’ in the ecosystem.  

When I think of it, any human social structure, such as settlements, industries, infrastructures etc., needs to stay in balance with natural environment. That balance is to be understood broadly, as the capacity to stay, for a satisfactorily long time, within a ‘safety zone’, where the ecosystem simply doesn’t kill us. That view has little to do with the moral concepts of environment-friendliness or sustainability. As a matter of fact, most known human social structures sooner or later fall out of balance with the ecosystem, and this is how civilizations collapse. Thus, here comes the first important assumption: any human social structure is, at some level, an environmental project. The incumbent social structures, possible to consider as relatively stable, are environmental projects which have simply hold in place long enough to grow social institutions, and those institutions allow further seeking of environmental balance.

I am starting my review of literature with an article by Phiri et al. (2021[1]), where the authors present a model for assessing the way that alluvial floodplains behave. I chose this one because my concept of Energy Ponds is supposed to work precisely in alluvial floodplains, i.e. in places where we have: a) a big river b) a lot of volatility in the amount of water in that river, and, as a consequence, we have (c) an alternation of floods and droughts. Normal stuff where I come from, i.e. in Northern Europe. Phiri et al. use the general model, acronymically called SWAT, which comes from ‘Soil and Water Assessment Tool’ (see also: Arnold et al. 1998[2]; Neitsch et al. 2005[3]), and with that general tool, they study the concept of pseudo-reservoirs in alluvial plains. In short, a pseudo-reservoir is a hydrological structure which works like a reservoir but does not necessarily look like one. In that sense, wetlands in floodplains can work as reservoirs of water, even if from the hydrological point of view they are rather extensions of the main river channel (Harvey et al. 2009[4]).

Analytically, the SWAT model defines the way a reservoir works with the following equation: V = Vstored + Vflowin − Vflowout + Vpcp − Vevap − Vseep . People can rightly argue that it is a good thing to know what symbols mean in an equation, and therefore V stands for the volume of water in reservoir at the end of the day, Vstored corresponds to the amount of water stored at the beginning of the day, Vflowin means the quantity of water entering reservoir during the day, Vflowout is the metric outflow of water during the day, Vpcp is volume of precipitation falling on the water body during the day, Vevap is volume of water removed from the water body by evaporation during the day, Vseep is volume of water lost from the water body by seepage.

This is a good thing to know, as well, once we have a nice equation, what the hell are we supposed to do with it in real life. Well, the SWAT model has even its fan page (http://www.swatusers.com ), and, as Phiri et al. phrase it out, it seems that the best practical use is to control the so-called ‘target release’, i.e. the quantity of water released at a given point in space and time, designated as Vtarg. The target release is mostly used as a control metric for preventing or alleviating floods, and with that purpose in mind, two decision rules are formulated. During the non-flood season, no reservation for flood is needed, and target storage is set at emergency spillway volume. In other words, in the absence of imminent flood, we can keep the reservoir full. On the other hand, when the flood season is on, flood control reservation is a function of soil water content. This is set to maximum and 50 % of maximum for wet and dry grounds, respectively. In the context of the V = Vstored + Vflowin − Vflowout + Vpcp − Vevap − Vseep equation, Vtarg is a specific value (or interval of values) in the Vflowout component.

As I am wrapping my mind around those conditions, I am thinking about the opposite application, i.e. about preventing and alleviating droughts. Drought is recognizable by exceptionally low values in the amount of water stored at the end of the given period, thus in the basic V, in the presence of low precipitation, thus low Vpcp, and high evaporation, which corresponds to high Vevap. More generally, both floods and droughts occur when – or rather after – in a given Vflowin − Vflowout balance, precipitation and evaporation take one turn or another.

I feel like moving those exogenous meteorological factors on one side of the equation, which goes like  – Vpcp + Vevap =  – V + Vstored + Vflowin − Vflowout − Vseep and doesn’t make much sense, as there are not really many cases of negative precipitation. I need to switch signs, and then it is more presentable, as Vpcp – VevapV – Vstored – Vflowin + Vflowout + Vseep . Weeell, almost makes sense. I guess that Vflowin is sort of exogenous, too. The inflow of water into the basin of the river comes from a melting glacier, from another river, from an upstream section of the same river etc. I reframe: Vpcp – Vevap + Vflowin V – Vstored + Vflowout + Vseep  . Now, it makes sense. Precipitations plus the inflow of water through the main channel of the river, minus evaporation, all that stuff creates a residual quantity of water. That residual quantity seeps into the groundwaters (Vseep), flows out (Vflowout), and stays in the reservoir-like structure at the end of the day (V – Vstored).

I am having a look at how Phiri et al. (2021 op. cit.) phrase out their model of pseudo-reservoir. The output value they peg the whole thing on is Vpsrc, or the quantity of water retained in the pseudo-reservoir at the end of the day. The Vpsrc is modelled for two alternative situations: no flood (V ≤ Vtarg), or flood (V > Vtarg). I interpret drought as particularly uncomfortable a case of the absence of flood.

Whatever. If V ≤ Vtarg , then Vpsrc = Vstored + Vflowin − Vbaseflowout + Vpcp − Vevap − Vseep  , where, besides the already known variables, Vbaseflowoutstands for volume of water leaving PSRC during the day as base flow. When, on the other hand, we have flood, Vpsrc = Vstored + Vflowin − Vbaseflowout − Voverflowout + Vpcp − Vevap − Vseep .

Phiri et al. (2021 op. cit.) argue that once we incorporate the phenomenon of pseudo-reservoirs in the evaluation of possible water discharge from alluvial floodplains, the above-presented equations perform better than the standard SWAT model, or V = Vstored + Vflowin − Vflowout + Vpcp − Vevap − Vseep

My principal takeaway from the research by Phiri et al. (2021 op. cit.) is that wetlands matter significantly for the hydrological balance of areas with characteristics of floodplains. My concept of ‘Energy Ponds’ assumes, among other things, storing water in swamp-like structures, including urban and semi-urban ones, such as rain gardens (Sharma & Malaviya 2021[5] ; Li, Liu & Li 2020[6] ; Venvik & Boogaard 2020[7],) or sponge cities (Ma, Jiang & Swallow 2020[8] ; Sun, Cheshmehzangi & Wang 2020[9]).  

Now, I have a few papers which allow me to have sort of a bird’s eye view of the SWAT model as regards the actual predictability of flow and retention in fluvial basins. It turns out that identifying optimal sites for hydropower installations is a very complex task, prone to a lot of error, and only the introduction of digital data such as GIS allows acceptable precision. The problem is to estimate accurately both the flow and the head of the waterway in question at an exact location (Liu et al., 2017[10]; Gollou and Ghadimi 2017[11]; Aghajani & Ghadimi 2018[12]; Yu & Ghadimi 2019[13]; Cai, Ye & Gholinia 2020[14]). My concept of ‘Energy Ponds’ includes hydrogeneration, but makes one of those variables constant, by introducing something like Roman siphons, with a constant head, apparently possible to peg at 20 metres. The hydro-power generation seems to be pseudo-concave function (i.e. it hits quite a broad, concave peak of performance) if the hydraulic head (height differential) is constant, and the associated productivity function is strongly increasing. Analytically, it can be expressed as a polynomial, i.e. as a combination of independent factors with various powers (various impact) assigned to them (Cordova et al. 2014[15]; Vieira et al. 2015[16]). In other words, by introducing, in my technological concept, that constant head (height) makes the whole thing more prone to optimization.

Now, I take on a paper which shows how to present a proof of concept properly: Pradhan, A., Marence, M., & Franca, M. J. (2021). The adoption of Seawater Pump Storage Hydropower Systems increases the share of renewable energy production in Small Island Developing States. Renewable Energy, https://doi.org/10.1016/j.renene.2021.05.151 . This paper is quite close to my concept of ‘Energy Ponds’, as it includes the technology of pumped storage, which I think about morphing and changing into something slightly different. Such as presented by Pradhan, Marence & Franca (2021, op. cit.), the proof of concept is structured in two parts: the general concept is presented, and then a specific location is studied  – the island of Curaçao, in this case – as representative for a whole category. The substance of proof is articulated around the following points:

>> the basic diagnosis as for the needs of the local community in terms of energy sources, with the basic question whether Seawater Pumped Storage Hydropower System is locally suitable as technology. In this specific case, the main criterium was the possible reduction of dependency on fossils. Assumptions as for the electric power required have been made, specifically for the local community.  

>> a GIS tool has been tested for choosing the optimal location. GIS stands for Geographic Information System (https://en.wikipedia.org/wiki/Geographic_information_system ). In this specific thread the proof of concept consisted in checking whether the available GIS data, and the software available for processing it are sufficient for selecting an optimal location in Curaçao.

At the bottom line, the proof of concept sums up to checking, whether the available GIS technology allows calibrating a site for installing the required electrical power in a Seawater Pumped Storage Hydropower System.

That paper by Pradhan, Marence & Franca (2021, op. cit.) presents a few other interesting traits for me. Firstly, the author’s prove that combining hydropower with windmills and solar modules is a viable solution, and this is exactly what I thought, only I wasn’t sure. Secondly, the authors consider a very practical issue: corrosion, and the materials recommended in order to bypass that problem. Their choice is fiberglass. Secondly, they introduce an important parameter, namely the L/H aka ‘Length to Head’ ratio. This is the proportion between the length of water conductors and the hydraulic head (i.e. the relative denivelation) in the actual installation. Pradhan, Marence & Franca recommend distinguishing two types of installations: those with L/H < 15, on the one hand, and those with 15 ≤ L/H ≤ 25. However accurate is that assessment of theirs, it is a paremeter to consider. In my concept of ‘Energy Ponds’, I assume an artificially created hydraulic head of 20 metres, and thus the conductors leading from elevated tanks to the collecting wetland-type structure should be classified in two types, namely [(L/H < 15) (L < 15*20) (L < 300 metres)], on the one hand, and [(15 ≤ L/H ≤ 25) (300 metres ≤ L ≤ 500 metres)], on the other hand.  

Still, there is bad news for me. According to a report by Botterud, Levin & Koritarov (2014[17]), which Pradhan, Marence & Franca quote as an authoritative source, hydraulic head for pumped storage should be at least 100 metres in order to make the whole thing profitable. My working assumption with ‘Energy Ponds’ is 20 metres, and, obviously, I have to work through it.

I think I have the outline of a structure for writing a decent proof-of-concept article for my ‘Energy Ponds’ concept. I think I should start with something I have already done once, two years ago, namely with compiling data as regards places in Europe, located in fluvial plains, with relatively the large volatility in water level and flow. These places will need water retention.

Out of that list, I select locations eligible for creating wetland-type structures for retaining water, either in the form of swamps, or as porous architectural structures. Once that second list prepared, I assess the local need for electrical power. From there, I reverse engineer. With a given power of X megawatts, I reverse to the storage capacity needed for delivering that power efficiently and cost-effectively. I nail down the storage capacity as such, and I pass in review the available technologies of power storage.

Next, I choose the best storage technology for that specific place, and I estimate the investment outlays necessary for installing it. I calculate the hydropower required in hydroelectric turbines, as well as in adjacent windmills and photovoltaic. I check whether the local river can supply the amount of water that fits the bill. I pass in review literature as regards optimal combinations of those three sources of energy. I calculate the investment outlays needed to install all that stuff, and I add the investment required in ram pumping, elevated tanks, and water conductors.  

Then, I do a first approximation of cash flow: cash from sales of electricity, in that local installation, minus the possible maintenance costs. After I calculate that gross margin of cash,  I compare it to the investment capital I had calculated before, and I try to estimate provisionally the time of return on investment. Once this done, I add maintenance costs to my sauce. I think that the best way of estimating these is to assume a given lifecycle of complete depreciation in the technology installed, and to count maintenance costs as the corresponding annual amortization.         


[1] Phiri, W. K., Vanzo, D., Banda, K., Nyirenda, E., & Nyambe, I. A. (2021). A pseudo-reservoir concept in SWAT model for the simulation of an alluvial floodplain in a complex tropical river system. Journal of Hydrology: Regional Studies, 33, 100770. https://doi.org/10.1016/j.ejrh.2020.100770.

[2] Arnold, J.G., Srinivasan, R., Muttiah, R.S., Williams, J.R., 1998. Large area hydrological modelling and assessment: Part I. Model development. J. Am. Water Resour. Assoc. 34, 73–89.

[3] Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R., 2005. “Soil and Water Assessment Tool Theoretical Documentation.” Version 2005. Blackland Research Center, Texas.

[4] Harvey, J.W., Schaffranek, R.W., Noe, G.B., Larsen, L.G., Nowacki, D.J., O’Connor, B.L., 2009. Hydroecological factors governing surface water flow on a low-gradient floodplain. Water Resour. Res. 45, W03421, https://doi.org/10.1029/2008WR007129.

[5] Sharma, R., & Malaviya, P. (2021). Management of stormwater pollution using green infrastructure: The role of rain gardens. Wiley Interdisciplinary Reviews: Water, 8(2), e1507. https://doi.org/10.1002/wat2.1507

[6] Li, J., Liu, F., & Li, Y. (2020). Simulation and design optimization of rain gardens via DRAINMOD and response surface methodology. Journal of Hydrology, 585, 124788. https://doi.org/10.1016/j.jhydrol.2020.124788

[7] Venvik, G., & Boogaard, F. C. (2020). Infiltration capacity of rain gardens using full-scale test method: effect of infiltration system on groundwater levels in Bergen, Norway. Land, 9(12), 520. https://doi.org/10.3390/land9120520

[8] Ma, Y., Jiang, Y., & Swallow, S. (2020). China’s sponge city development for urban water resilience and sustainability: A policy discussion. Science of the Total Environment, 729, 139078. https://doi.org/10.1016/j.scitotenv.2020.139078

[9] Sun, J., Cheshmehzangi, A., & Wang, S. (2020). Green infrastructure practice and a sustainability key performance indicators framework for neighbourhood-level construction of sponge city programme. Journal of Environmental Protection, 11(2), 82-109. https://doi.org/10.4236/jep.2020.112007

[10] Liu, Yan, Wang, Wei, Ghadimi, Noradin, 2017. Electricity load forecasting by an improved forecast engine for building level consumers. Energy 139, 18–30. https://doi.org/10.1016/j.energy.2017.07.150

[11] Gollou, Abbas Rahimi, Ghadimi, Noradin, 2017. A new feature selection and hybrid forecast engine for day-ahead price forecasting of electricity markets. J. Intell. Fuzzy Systems 32 (6), 4031–4045.

[12] Aghajani, Gholamreza, Ghadimi, Noradin, 2018. Multi-objective energy manage- ment in a micro-grid. Energy Rep. 4, 218–225.

[13] Yu, Dongmin, Ghadimi, Noradin, 2019. Reliability constraint stochastic UC by considering the correlation of random variables with Copula theory. IET Renew. Power Gener. 13 (14), 2587–2593.

[14] Cai, X., Ye, F., & Gholinia, F. (2020). Application of artificial neural network and Soil and Water Assessment Tools in evaluating power generation of small hydropower stations. Energy Reports, 6, 2106-2118. https://doi.org/10.1016/j.egyr.2020.08.010.

[15] Cordova M, Finardi E, Ribas F, de Matos V, Scuzziato M. Performance evaluation and energy production optimization in the real-time operation of hydropower plants. Electr Pow Syst Res 2014;116:201–7.   http://dx.doi.org/ 10.1016/j.epsr.2014.06.012  

[16] Vieira, D. A. G., Guedes, L. S. M., Lisboa, A. C., & Saldanha, R. R. (2015). Formulations for hydroelectric energy production with optimality conditions. Energy Conversion and Management, 89, 781-788.

[17] Botterud, A., Levin, T., & Koritarov, V. (2014). Pumped storage hydropower: benefits for grid reliability and integration of variable renewable energy (No. ANL/DIS-14/10). Argonne National Lab.(ANL), Argonne, IL (United States). https://publications.anl.gov/anlpubs/2014/12/106380.pdf

When it plays out, it looks nasty

I feel like using my hypothesis of collectively intelligent social structures in other fields than just energy and urbanisation, which I have been largely doing so far. This time, I want to make a case for individual freedom as both a factor and a manifestation of collective intelligence. There is a population of humans. Each human has m possible states of being. As soon as two humans interact, one m states of being in the first human interacts with the other m states of being in the other human. It is like an existential geometrical square: those two humans together have m*m = m2 collective states of being. Generally, n humans, with m possible states of being in each of them, can produce mn different states of being together. When n gets substantial, like 38 million people in my home country, Poland, you can hardly expect all of us 38 million Poles having the repertoire of freedom in our behavioural patterns. Some of us will have 3m actually happening states of being, some other will soar into 6m alternative ways of being in the world, whilst still some other others will modestly stick to 0,3m. In that large population, the standard m ways of existing will be an expected state, thus an arithmetical average or an expected interval around it.   

Collectively intelligent structures learn by experimenting with many alternative states of themselves. Up to a point, the more such alternative states, the more and better we can learn. There is probably a point where ‘the more’ becomes ‘too much to process’, and then, we face a fork on the road: either we simply ignore some alternative versions of ourselves and we truly learn just from those which we can cover inside our cognitive span, or we try to experiment with everything we can possibly be, and chaos develops. I understand freedom, at the collective level, as the flexibility in shifting between those different states of being. Organized, collective freedom is the ability to explore the sweet spot of transition between order and chaos, and the ability to experiment with as many alternative versions of ourselves as we possibly can. Those collectively defined alternative realities always follow the basic logic of mn. At the end of the day, there are as many versions of us being together as there are us, for one, namely the ‘n’ exponent, and as many as there are possible states of being in the average individual among n, and this is the ‘m’ base.

Degrees of freedom in the average member of society are the foundation of collectively intelligent learning. I guess this is a mathematical argument for individual freedom in legal and political systems. As I think about my whole hypothesis of collectively intelligent social structures, I inevitably ask the question which any social scientist needs to ask: what is the practical usefulness of all that stuff? Social sciences are applied sciences, at the end of the day. However abstract I go in my intellectual peregrinations, my findings and methods need to serve in real life, for designing policies, business strategies, business plans etc. The empirical method I have developed around that whole thing of collective intelligence opens on two practical applications. Firstly, it allows non-arbitrary testing of various empirical observables as actual social outcomes. In policies and business strategies, and, by the way, in the whole realm of social sciences, there is that curse of arbitrary orientations. ‘People strive to maximize profit’. ‘No, they want to optimize dynamic equilibriums in their social games’. ‘Well, maybe, but we can and should educate people towards social justice and environmentally rational behaviour’ etc. etc. All that chatter abounds in literature which deems itself ‘scientific’, and yet it is 100% metaphysics, with no scientific grounds at all. I think my method allows working around that metaphysical part and testing human populations for the actual outcomes they collectively, objectively pursue. Here comes an interesting question: are our goals collective or individual? The more I think about it, the more I am convinced they are collective. When I ask myself about my own goals, at least those which I phrase out explicitly in my mind, they are all sort of categorical rather than idiosyncratically my own. I pursue the types of goals which many other people pursue in their existence. I just hop on those specific wagons, with my own backpack.  

Secondly, my method allows exploring the issue of Black Swans, i.e. outlier events, which suddenly become key drivers of social change. The method I have developed allows simulating something like a social chain reaction. An unexpected triggering event happens, and it is unexpected because from our point of view it is random. That triggers a collection of events which we could otherwise fathom, but they have been in the refrigerator of history so far. Now, they are triggered into existence, and, at the same time, the overall cohesion of the social structure weakens, at least temporarily. New things start happening, and old things happen sort of more loosely and chaotically than they used to. I have discovered that depending on the exact orientation assigned a priori to the social structure I study, those social chain reactions can we essentially predictable, completely unpredictable, or, in still another case, we can calm them down exaggeratedly quickly, without really learning from them.

All in all, the method of using a simple neural network as social simulator, which I developed in connection with my hypothesis of collectively intelligent social structures, allows what I perceive as very empiricist a study of social change, much freer of metaphysics than many other methods. Of course, a bit of metaphysics is unavoidable. What we use to call ‘quantitative variables’ in social sciences are always the mathematics of something we think that happens, and we think in terms of our language and culture.

Ooops, pardon my manners, I have gone into philosophy again. Philosophy is nice, but when I stay in this realm longer than what is strictly necessary for feeling like an intellectual, I start feeling as too much of an intellectual and my apish side calls for more ground under my feet. I use this blog for providing a current account of my intellectual journey, and of the actual projects which I am working on. I hope that the paragraphs above are (provisionally) sufficient as regards the intellectual journey, and I can pass to debriefing on my projects.

One of the projects I start working on is a platform for debt-based crowdfunding. This is some sort of comeback to the interest I had in financial schemes for the implementation of small installations in renewable energies. For the less initiated readers, I am quickly going through the basics. You probably know that if your cousin asks you to invest in his or her business, you can do it, on the basis of a private contract of partnership, and, in most countries, you don’t even go to jail afterwards. This is the market of private equity. You can also lend money to your cousin, you can agree as for the exact terms of the loan, and this is financing through private debt. The opposite of private is public, and therefore we have public capital markets on the opposite end of the spectrum. Stock markets are the most visible ones, and sort of next to them are the markets of publicly traded debt, where you can buy and sell bonds of all kinds: corporate, municipal, and sovereign. 

Between the strictly private and the regulated public, a transitional zone, of many shades and colours, is to be found.  Crowdfunding, sometimes called ‘societal funding’ or ‘communitarian funding’ dwells in this zone, precisely. The basic difference between crowdfunding and private finance strictly spoken is the largely aleatory, social-media-type creation of relations between investors and entrepreneurs. Crowdfunding happens essentially via digital platforms, where entrepreneurs auction their ventures and try to attract whoever is interested in them. Those digital platforms in themselves are marketing engines, essentially. On the other hand, the basic difference between public financial markets and crowdfunding is that the latter does not really allow tradability in financial positions. When I invest my money through crowdfunding, it is much more of a long-term commitment than investment via stock market. Less liquidity in my financial assets means more exposure to long-term risks, and yet less exposure to short-term volatility in market value.

In my own big picture of social reality, I put the emergence of crowdfunding in the same phenomenological bag as I put cryptocurrencies, progressively increasing supply of money in relation to real output in the economy (thus decreasing velocity of money), and increasingly cash-furnished corporate balance sheets. As a civilisation, we are building up a growing base of financial liquidity, and that means we are facing a quickening pace of depreciation in technological assets, and thus we are in the middle of accelerated technological change. Now, a little word is due about the way I understand accelerated technological change. I have encountered quite well-articulated views that technological change is currently disappointingly slow as compared to what we need. Well, maybe, but in strictly spoken business terms, when a piece of technology which I purchased last year ages morally twice as fast as those which I purchased 5 years ago, because new generations of the same equipment pop up faster and faster, this is accelerated technological change, and, as a businessperson, I need to figure out a strategy to cope with that change.

Here, my own point of view of that phenomenon called ‘financialization’ differs significantly from a lot of other researchers. The mainstream doctrine says that increased financialization is a bad thing, it destabilizes the economic system, and it contributes to social inequalities. I think that financialization is the by-product of something else. It is an otherwise rational coping mechanism to smooth and amortize quick social change which, without financialization, could take very nasty forms, like global wars, massive disappearance of human settlements and much greater damage to natural environment than what we use to bitch and moan about today. Just imagine that somewhere in Europe, 5 million people in a post-industrial spot cannot afford to pay for electricity anymore and they start burning wood and coal in stoves instead. This is what could happen in the presence of quick technological change and in the absence of that horrible financialization.     

Crowdfunding is essentially attached to new ideas and new business structures. It is seed capital or early development capital. When I invest my money through crowdfunding, I am opening a long-term position in something essentially young, burgeoning and full of uncertainty. One hundred years ago, mustering capital for such a venture would take an entrepreneur years of patient contacts with potential investors. Now, it can take months or even weeks, and this is the tangible gain of time through the use of digital platforms.   

That introduction kept in mind, I get closer to the main thread of that project in crowdfunding, namely to the new regulations thereof, likely to enter into force in Poland this autumn, based on recent regulations of the European Union as a whole. I am passing in review the REGULATION (EU) 2020/1503 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 7 October 2020 on European crowdfunding service providers for business, and amending Regulation (EU) 2017/1129 and Directive (EU) 2019/1937, to find at https://eur-lex.europa.eu/legal-content/PL/TXT/?uri=CELEX:32020R1503 . As I usually do, I start from the end, more specifically from Annex II, titled SOPHISTICATED INVESTORS FOR THE PURPOSE OF THIS REGULATION.

A sophisticated investor is an investor who possesses the awareness of the risks associated with investing in capital markets and adequate resources to undertake those risks without exposing itself to excessive financial consequences. Sophisticated investors may be categorised as such if they meet identification criteria, which, in turn, differ according to the legal personality of the entity. Legal persons (like a bunch of folks in a business partnership), are assumed to be sophisticated in their investments if they meet at least one of the following criteria: (a) own funds of at least EUR 100 000 (b) net turnover of at least EUR 2 000 000 (c) balance sheet of at least EUR 1 000 000.

On the other hand, natural persons can call themselves sophisticated investors when the meet at least two of the following criteria:

>> (a) personal gross income of at least EUR 60 000 per fiscal year, or a financial instrument portfolio, defined as including cash deposits and financial assets, that exceeds EUR 100 000;

>> (b) the investor works or has worked in the financial sector for at least one year in a professional position which requires knowledge of the transactions or services envisaged, or the investor has held an executive position for at least 12 months in a legal person considered as sophisticated investor;

>> (c) the investor has carried out transactions of a significant size on the capital markets at an average frequency of 10 per quarter, over the previous four quarters.

The whole distinction between ordinary investors and the sophisticated ones is in the degree of legal protection they are provided with. That distinction essentially taps into an older one, contained in the DIRECTIVE 2014/65/EU OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 15 May 2014 on markets in financial instruments and amending Directive 2002/92/EC and Directive 2011/61/EU (https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A32014L0065 ). As it happens sometimes, protection turns out to be a limitation actually. Non-sophisticated investors are generally limited in the amounts of money they can invest, and the repertoire of financial instruments which they can invest in. If one wants not to be treated like a child, they have to make a special, written request to be treated as sophisticated investor, and whatever operator of financial platform is that request addressed to can accept or reject said request.    

The Polish prospective regulations on crowdfunding approach things from a different angle. By the way, they are just prospective regulations, and the only official version of that will which I could get my hands on is in Polish. For those who speak the beautiful language of my home country – distinctive, among others, by a record-level density of consonants in one word – I placed the current bill of this regulation in the archives of my blog, just here: https://discoversocialsciences.com/wp-content/uploads/2021/05/Projekt-crowdfunding-.docx . Polish regulators focus mostly on the concept of ‘key investment information sheet’, which I will allow myself to call KIIS in what follows, is present in the European regulations as well. The KIIS should warn prospective investors that the investing environment they have entered into entails risks that are covered neither by deposit guarantee schemes, nor by investor compensation schemes. The KIIS should reflect the specific features of lending-based and investment-based crowdfunding. To that end, specific and relevant indicators should be required. The KIIS should also take into account, where available, the specific features and risks associated with project owners, and should focus on material information about the project owners, the investors’ rights and fees, and the type of transferable securities, admitted instruments for crowdfunding purposes and loans offered. The KIIS should be drawn up by the project owners, because the project owners are in the best position to provide the information required to be included therein. However, since it is the crowdfunding service providers that are responsible for providing the KIIS to prospective investors, it is the crowd­funding service providers that should ensure that the KIIS is clear, correct and complete.

The specificity of the Polish regulations as regards the KIIS is largely in the addressees of that information. In the general European regulations, the KIIS is addressed to prospective and actual investors. In Polish regulations, it is strongly stressed that crowdfunding operators should communicate all their KIIS’s to the Financial Supervision Commission (PL: Komisja Nadzoru Finansowego, https://www.knf.gov.pl/en/ ), not later than 7 days before making the same KIIS available to prospective investors. On the other hand, the owner of the project subject to crowdfunding can publish the KIIS on their own platform only after the provider of crowdfunding does in on their own one. We have a sequence of KIISes. The first KIIS goes from the crowdfunding provider to the Financial Supervision Commission, which has at least 7 days to consider (what exactly?). The next KIIS goes from the crowdfunding provider to prospective investors, who also receive the last KIIS from the owner of the crowdfunded project in question.

In a general manner, those Polish regulations give a lot of discretionary prerogatives to the Financial Supervision Commission as regards crowdfunding providers. They can halt a crowdfunding project immediately, and for an essentially indefinite period of time, on the grounds of a simple suspicion. I don’t like it. Someone in charge with the Financial Supervision Commission is the first to know about a crowdfunded project, they can request any information about that project, they can halt the project whenever they want. That smells bad. That smells insider trading. That smells uncontrolled pressure on the owners of crowdfunded projects. Imagine: you start such a project, and then you have a phone call, I mean THE phone call. Someone tells you they know about your crowdfunding campaign, and they would willingly take 60% of your business for 50% of its book value. You refuse, and the next thing you know is your crowdfunding campaign being suspended for an unknown period of time. I know the scheme, I saw it play out, and when it plays out, it looks nasty, believe me. That means people close to the government taking over entire swaths of small business, and the kind of small business, which is particularly exposed to adverse actions, the emerging one.  

Investment, national security, and psychiatry

I need to clear my mind a bit. For the last few weeks, I have been working a lot on revising an article of mine, and I feel I need a little bit of a shake-off. I know by experience that I need a structure to break free from another structure. Yes, I am one of those guys. I like structures. When I feel I lack one, I make one.

The structure which I want to dive into, in order to shake off the thinking about my article, is the thinking about my investment in the stock market. My general strategy in that department is to take the rent, which I collect from an apartment in town, every month, and to invest it in the stock market. Economically, it is a complex process of converting the residential utility of a real asset (apartment) into a flow of cash, thus into a financial asset with quite steady a market value (inflation is still quite low), and then I convert that low-risk financial asset into a differentiated portfolio of other financial assets endowed with higher a risk (stock). I progressively move capital from markets with low risk (residential real estate, money) into a high-risk-high-reward market.

I am playing a game. I make a move (monthly cash investment), and I wait for a change in the stock market. I am wrapping my mind around the observable change, and I make my next move the next month. With each move I make, I gather information. What is that information? Let’s have a look at my portfolio such as it is now. You can see it in the table below:

StockValue in EURReal return in €Rate of return I have as of April 6ht, 2021, in the morning
CASH & CASH FUND & FTX CASH (EUR) € 25,82 €                                    –   €                                     25,82
ALLEGRO.EU SA € 48,86 €                               (2,82)-5,78%
ALTIMMUNE INC. – COMM € 1 147,22 €                            179,6515,66%
APPLE INC. – COMMON ST € 1 065,87 €                                8,210,77%
BIONTECH SE € 1 712,88 €                           (149,36)-8,72%
CUREVAC N.V. € 711,00 €                             (98,05)-13,79%
DEEPMATTER GROUP PLC € 8,57 €                               (1,99)-23,26%
FEDEX CORPORATION COMM € 238,38 €                              33,4914,05%
FIRST SOLAR INC. – CO € 140,74 €                             (11,41)-8,11%
GRITSTONE ONCOLOGY INC € 513,55 €                           (158,43)-30,85%
INPOST € 90,74 €                             (17,56)-19,35%
MODERNA INC. – COMMON € 879,85 €                             (45,75)-5,20%
NOVAVAX INC. – COMMON STOCK € 1 200,75 €                            398,5333,19%
NVIDIA CORPORATION – C € 947,35 €                              42,254,46%
ONCOLYTICS BIOTCH CM € 243,50 €                             (14,63)-6,01%
SOLAREDGE TECHNOLOGIES € 683,13 €                             (83,96)-12,29%
SOLIGENIX INC. COMMON € 518,37 €                           (169,40)-32,68%
TESLA MOTORS INC. – C € 4 680,34 €                            902,3719,28%
VITALHUB CORP.. € 136,80 €                               (3,50)-2,56%
WHIRLPOOL CORPORATION € 197,69 €                              33,1116,75%
  €       15 191,41 €                            840,745,53%

A few words of explanation are due. Whilst I have been actively investing for 13 months, I made this portfolio in November 2020, when I did some major reshuffling. My overall return on the cash invested, over the entire period of 13 months, is 30,64% as for now (April 6th, 2021), which makes 30,64% * (12/13) = 28,3% on the annual basis.

The 5,53% of return which I have on this specific portfolio makes roughly 1/6th of the total return in have on all the portfolios I had over the past 13 months. It is the outcome of my latest experimental round, and this round is very illustrative of the mistake which I know I can make as an investor: panic.

In August and September 2020, I collected some information, I did some thinking, and I made a portfolio of biotech companies involved in the COVID-vaccine story: Pfizer, Biontech, Curevac, Moderna, Novavax, Soligenix. By mid-October 2020, I was literally swimming in extasy, as I had returns on these ones like +50%. Pure madness. Then, big financial sharks, commonly called ‘investment funds’, went hunting for those stocks, and they did what sharks do: they made their target bleed before eating it. They boxed and shorted those stocks in order to make their prices affordably low for long investment positions. At the time, I lost control of my emotions, and when I saw those prices plummet, I sold out everything I had. Almost as soon as I did it, I realized what an idiot I had been. Two weeks later, the same stocks started to rise again. Sharks had had their meal. In response, I did what I still wonder whether it was wise or stupid: I bought back into those positions, only at a price higher than what I sold them for.

Selling out was stupid, for sure. Was buying back in a wise move? I don’t know, like really. My intuition tells me that biotech companies in general have a bright future ahead, and not only in connection with vaccines. I am deeply convinced that the pandemic has already built up, and will keep building up an interest for biotechnology and medical technologies, especially in highly innovative forms. This is even more probable as we realized that modern biotechnology is very largely digital technology. This is what is called ‘platforms’ in the biotech lingo. These are digital clouds which combine empirical experimental data with artificial intelligence, and the latter is supposed to experiment virtually with that data. Modern biotechnology consists in creating as many alternative combinations of molecules and lifeforms as we possibly can make and study, and then pick those which offer the best combination of biological outcomes with the probability of achieving said outcomes.

My currently achieved rates of return, in the portfolio I have now, are very illustrative of an old principle in capital investment: I will fail most of the times. Most of my investment decisions will be failures, at least in the short and medium term, because I cannot possibly outsmart the incredibly intelligent collective structure of the stock market. My overall gain, those 5,53% in the case of this specific portfolio, is the outcome of 19 experiments, where I fail in 12 of them, for now, and I am more or less successful in the remaining 7.

The very concept of ‘beating the market’, which some wannabe investment gurus present, is ridiculous. The stock market is made of dozens of thousands of human brains, operating in correlated coupling, and leveraged with increasingly powerful artificial neural networks. When I expect to beat that networked collective intelligence with that individual mind of mine, I am pumping smoke up my ass. On the other hand, what I can do is to do as many different experiments as I can possibly spread my capital between.

It is important to understand that any investment strategy, where I assume that from now on, I will not make any mistakes, is delusional. I made mistakes in the past, and I am likely to make mistakes in the future. What I can do is to make myself more predictable to myself. I can narrow down the type of mistakes I tend to make, and to create the corresponding compensatory moves in my own strategy.

Differentiation of risk is a big principle in my investment philosophy, and yet it is not the only one. Generally, with the exception of maybe 2 or 3 days in a year, I don’t really like quick, daily trade in the stock market. I am more of a financial farmer: I sow, and I wait to see plants growing out of those seeds. I invest in industries rather than individual companies. I look for some kind of strong economic undertow for my investments, and the kind of undertow I specifically look for is high potential for deep technological change. Accessorily, I look for industries which sort of logically follow human needs, e.g. the industry of express deliveries in the times of pandemic. I focus on three main fields of technology: biotech, digital, and energy.

Good. I needed to shake off, and I am. Thinking and writing about real business decisions helped me to take some perspective. Now, I am gently returning into the realm of science, without completely leaving the realm of business: I am navigating the somehow troubled and feebly charted waters of money for science. I am currently involved in launching and fundraising for two scientific projects, in two very different fields of science: national security and psychiatry. Yes, I know, they can conjunct in more points than we commonly think they can. Still, in canonical scientific terms, these two diverge.

How come I am involved, as researcher, in both national security and psychiatry? Here is the thing: my method of using a simple artificial neural network to simulate social interactions seems to be catching on. Honestly, I think it is catching on because other researchers, when they hear me talking about ‘you know, simulating alternative realities and assessing which one is the closest to the actual reality’ sense in me that peculiar mental state, close to the edge of insanity, but not quite over that edge, just enough to give some nerve and some fun to science.

In the field of national security, I teamed up with a scientist strongly involved in it, and we take on studying the way our Polish forces of Territorial Defence have been acting in and coping with the pandemic of COVID-19. First, the context. So far, the pandemic has worked as a magnifying glass for all the f**kery in public governance. We could all see a minister saying ‘A,B and C will happen because we said so’, and right after there was just A happening, with a lot of delay, and then a completely unexpected phenomenal D appeared, with B and C bitching and moaning they haven’t the right conditions for happening decently, and therefore they will not happen at all.  This is the first piece of the context. The second is the official mission and the reputation of our Territorial Defence Forces AKA TDF. This is a branch of our Polish military, created in 2017 by our right-wing government. From the beginning, these guys had the reputation to be a right-wing militia dressed in uniforms and paid with taxpayers’ money. I honestly admit I used to share that view. TDF is something like the National Guard in US. These are units made of soldiers who serve in the military, and have basic military training, but they have normal civilian lives besides. They have civilian jobs, whilst training regularly and being at the ready should the nation call.

The initial idea of TDF emerged after the Russian invasion of the Crimea, when we became acutely aware that military troops in nondescript uniforms, apparently lost, and yet strangely connected to the Russian government, could massively start looking lost by our Eastern border. The initial idea behind TDF was to significantly increase the capacity of the Polish population for mobilising military resources. Switzerland and Finland largely served as models.

When the pandemic hit, our government could barely pretend they control the situation. Hospitals designated as COVID-specific had frequently no resources to carry out that mission. Our government had the idea of mobilising TDF to help with basic stuff: logistics, triage and support in hospitals etc. Once again, the initial reaction of the general public was to put the label of ‘militarisation’ on that decision, and, once again, I was initially thinking this way. Still, some friends of mine, strongly involved as social workers supporting healthcare professionals, started telling me that working with TDF, in local communities, was nothing short of amazing. TDF had the speed, the diligence, and the capacity to keep their s**t together which many public officials lacked. They were just doing their job and helping tremendously.

I started scratching the surface. I did some research, and I found out that TDF was of invaluable help for many local communities, especially outside of big cities. Recently, I accidentally had a conversation about it with M., the scientist whom I am working with on that project. He just confirmed my initial observations.

M. has strong connections with TDF, including their top command. Our common idea is to collect abundant, interview-based data from TDF soldiers mobilised during the pandemic, as regards the way they carried out their respective missions. The purely empirical edge we want to have here is oriented on defining successes and failures, as well as their context and contributing factors. The first layer of our study is supposed to provide the command of TDF with some sort of case-studies-based manual for future interventions. At the theoretical, more scientific level, we intend to check the following hypotheses:      

>> Hypothesis #1: during the pandemic, TDF has changed its role, under the pressure of external events, from the initially assumed, properly spoken territorial defence, to civil defence and assistance to the civilian sector.

>> Hypothesis #2: the actual role played by the TDF during the pandemic was determined by the TDF’s actual capacity of reaction, i.e. speed and diligence in the mobilisation of human and material resources.

>> Hypothesis #3: collectively intelligent human social structures form mechanisms of reaction to external stressors, and the chief orientation of those mechanisms is to assure proper behavioural coupling between the action of external stressors, and the coordinated social reaction. Note: I define behavioural coupling in terms of the games’ theory, i.e. as the objectively existing need for proper pacing in action and reaction.   

The basic method of verifying those hypotheses consists, in the first place, in translating the primary empirical material into a matrix of probabilities. There is a finite catalogue of operational procedures that TDF can perform. Some of those procedures are associated with territorial military defence as such, whilst other procedures belong to the realm of civil defence. It is supposed to go like: ‘At the moment T, in the location A, procedure of type Si had a P(T,A, Si) probability of happening’. In that general spirit, Hypothesis #1 can be translated straight into a matrix of probabilities, and phrased out as ‘during the pandemic, the probability of TDF units acting as civil defence was higher than seeing them operate as strict territorial defence’.

That general probability can be split into local ones, e.g. region-specific. On the other hand, I intuitively associate Hypotheses #2 and #3 with the method which I call ‘study of orientation’. I take the matrix of probabilities defined for the purposes of Hypothesis #1, and I put it back to back with a matrix of quantitative data relative to the speed and diligence in action, as regards TDF on the one hand, and other public services on the other hand. It is about the availability of vehicles, capacity of mobilisation in people etc. In general, it is about the so-called ‘operational readiness’, which you can read more in, for example, the publications of RAND Corporation (https://www.rand.org/topics/operational-readiness.html).  

Thus, I take the matrix of variables relative to operational readiness observable in the TDF, and I use that matrix as input for a simple neural network, where the aggregate neural activation based on those metrics, e.g. through a hyperbolic tangent, is supposed to approximate a specific probability relative to TDF people endorsing, in their operational procedures, the role of civil defence, against that of military territorial defence. I hypothesise that operational readiness in TDF manifests a collective intelligence at work and doing its best to endorse specific roles and applying specific operational procedures. I make as many such neural networks as there are operational procedures observed for the purposes of Hypothesis #1. Each of these networks is supposed to represent the collective intelligence of TDF attempting to optimize, through its operational readiness, the endorsement and fulfilment of a specific role. In other words, each network represents an orientation.

Each such network transforms the input data it works with. This is what neural networks do: they experiment with many alternative versions of themselves. Each experimental round, in this case, consists in a vector of metrics informative about the operational readiness TDF, and that vector locally tries to generate an aggregate outcome – its neural activation – as close as possible to the probability of effectively playing a specific role. This is always a failure: the neural activation of operational readiness always falls short of nailing down exactly the probability it attempts to optimize. There is always a local residual error to account for, and the way a neural network (well, my neural network) accounts for errors consists in measuring them and feeding them into the next experimental round. The point is that each such distinct neural network, oriented on optimizing the probability of Territorial Defence Forces endorsing and fulfilling a specific social role, is a transformation of the original, empirical dataset informative about the TDF’s operational readiness.

Thus, in this method, I create as many transformations (AKA alternative versions) of the actual operational readiness in TDF, as there are social roles to endorse and fulfil by TDF. In the next step, I estimate two mathematical attributes of each such transformation: its Euclidean distance from the original empirical dataset, and the distribution of its residual error. The former is informative about similarity between the actual reality of TDF’s operational readiness, on the one hand, and alternative realities, where TDF orient themselves on endorsing and fulfilling just one specific role. The latter shows the process of learning which happens in each such alternative reality.

I make a few methodological hypotheses at this point. Firstly, I expect a few, like 1 ÷ 3 transformations (alternative realities) to fall particularly close from the actual empirical reality, as compared to others. Particularly close means their Euclidean distances from the original dataset will be at least one order of magnitude smaller than those observable in the remaining transformations. Secondly, I expect those transformations to display a specific pattern of learning, where the residual error swings in a predictable cycle, over a relatively wide amplitude, yet inside that amplitude. This is a cycle where the collective intelligence of Territorial Defence Forces goes like: ‘We optimize, we optimize, it goes well, we narrow down the error, f**k!, we failed, our error increased, and yet we keep trying, we optimize, we optimize, we narrow down the error once again…’ etc. Thirdly, I expect the remaining transformations, namely those much less similar to the actual reality in Euclidean terms, to display different patterns of learning, either completely dishevelled, with the residual error bouncing haphazardly all over the place, or exaggeratedly tight, with error being narrowed down very quickly and small ever since.

That’s the outline of research which I am engaging into in the field of national security. My role in this project is that of a methodologist. I am supposed to design the system of interviews with TDF people, the way of formalizing the resulting data, binding it with other sources of information, and finally carrying out the quantitative analysis. I think I can use the experience I already have with using artificial neural networks as simulators of social reality, mostly in defining said reality as a vector of probabilities attached to specific events and behavioural patterns.     

As regards psychiatry, I have just started to work with a group of psychiatrists who have abundant professional experience in two specific applications of natural language in the diagnosing and treating psychoses. The first one consists in interpreting patients’ elocutions as informative about their likelihood of being psychotic, relapsing into psychosis after therapy, or getting durably better after such therapy. In psychiatry, the durability of therapeutic outcomes is a big thing, as I have already learnt when preparing for this project. The second application is the analysis of patients’ emails. Those psychiatrists I am starting to work with use a therapeutic method which engages the patient to maintain contact with the therapist by writing emails. Patients describe, quite freely and casually, their mental state together with their general existential context (job, family, relationships, hobbies etc.). They don’t necessarily discuss those emails in subsequent therapeutic sessions; sometimes they do, sometimes they don’t. The most important therapeutic outcome seems to be derived from the very fact of writing and emailing.

In terms of empirical research, the semantic material we are supposed to work with in that project are two big sets of written elocutions: patients’ emails, on the one hand, and transcripts of standardized 5-minute therapeutic interviews, on the other hand. Each elocution is a complex grammatical structure in itself. The semantic material is supposed to be cross-checked with neurological biomarkers in the same patients. The way I intend to use neural networks in this case is slightly different from that national security thing. I am thinking about defining categories, i.e. about networks which guess similarities and classification out of crude empirical data. For now, I make two working hypotheses:

>> Hypothesis #1: the probability of occurrence in specific grammatical structures A, B, C, in the general grammatical structure of a patient’s elocutions, both written and spoken, is informative about the patient’s mental state, including the likelihood of psychosis and its specific form.

>> Hypothesis #2: the action of written self-reporting, e.g. via email, from the part of a psychotic patient, allows post-clinical treatment of psychosis, with results observable as transition from mental state A to mental state B.

The right side of the disruption

I am swivelling my intellectual crosshairs around, as there is a lot going on, in the world. Well, there is usually a lot going on, in the world, and I think it is just the focus of my personal attention that changes its scope. Sometimes, I pay attention just to the stuff immediately in front of me, whilst on other times I go wide and broad in my perspective.

My research on collective intelligence, and on the application of artificial neural networks as simulators thereof has brought me recently to studying outlier cases. I am an economist, and I do business in the stock market, and therefore it comes as sort of logical that I am interested in business outliers. I hold some stock of the two so-far winners of the vaccine race: Moderna (https://investors.modernatx.com/ ) and BionTech (https://investors.biontech.de/investors-media ), the vaccine companies. I am interested in the otherwise classical, Schumpeterian questions: to what extent are their respective business models predictors of their so-far success in the vaccine contest, and, seen from the opposite perspective, to what extent is that whole technological race of vaccines predictive of the business models which its contenders adopt?

I like approaching business models with the attitude of a mean detective. I assume that people usually lie, and it starts with lying to themselves, and that, consequently, those nicely rounded statements in annual reports about ‘efficient strategies’ and ‘ambitious goals’ are always bullshit to some extent. In the same spirit, I assume that I am prone to lying to myself. All in all, I like falling back onto hard numbers, in the first place. When I want to figure out someone’s business model with a minimum of preconceived ideas, I start with their balance sheet, to see their capital base and the way they finance it, just to continue with their cash-flow. The latter helps my understanding on how they make money, at the end of the day, or how they fail to make any.

I take two points in time: the end of 2019, thus the starting blocks of the vaccine race, and then the latest reported period, namely the 3rd quarter of 2020. Landscape #1: end of 2019. BionTech sports $885 388 000 in total assets, whilst Moderna has $1 589 422 000. Here, a pretty amazing detail pops up. I do a routine check of proportion between fixed assets and total assets. It is about to see what percentage of the company’s capital base is immobilized, and thus supposed to bring steady capital returns, as opposed to the current assets, fluid, quick to exchange and made for greasing the current working of the business. When I measure that coefficient ‘fixed assets divided by total assets’, it comes as 29,8% for BionTech, and 29% for Moderna. Coincidence? There is a lot of coincidence in those two companies. When I switch to Landscape #2: end of September 2020, it is pretty much the. You can see it in the two tables below:

As you look at those numbers, they sort of collide with the common image of biotech companies in sci fi movies. In movies, we can see huge labs, like 10 storeys underground, with caged animals inside etc. In real life, biotech is cash, most of all. Biotech companies are like big wallets, camped next to some useful science. Direct investment in biotech means very largely depositing one’s cash on the bank account run by the biotech company.

After studying the active side of those two balance sheets, i.e. in BionTech and in Moderna, I shift my focus to the passive side. I want to know how exactly people put cash in those businesses. I can see that most of it comes in the form of additional paid-in equity, which is an interesting thing for publicly listed companies. In the case of Moderna, the bulk of that addition to equity comes as a mechanism called ‘vesting of restricted common stock’. Although it is not specified in their financial report how exactly that vesting takes place, the generic category corresponds to operations where people close to the company, employees or close collaborators, anyway in a closed private circle, buy stock of the company in a restricted issuance.  With Biontech, it is slightly different. Most of the proceeds from public issuance of common stock is considered as reserve capital, distinct from share capital, and on the top of that they seem to be running, similarly to Moderna, transactions of vesting restricted stock. Another important source of financing in both companies are short-term liabilities, mostly deferred transactional payments. Still, I have an intuitive impression of being surrounded by maybies (you know: ‘maybe I am correct, unless I am wrong), and thus I decided to broaden my view. I take all the 7 biotech companies I currently have in my investment portfolio, which are, besides BionTech and Moderna, five others: Soligenix (http://ir.soligenix.com/ ), Altimmune (http://ir.altimmune.com/investors ), Novavax (https://ir.novavax.com/ ) and VBI Vaccines (https://www.vbivaccines.com/investors/  ). In the two tables below, I am trying to summarize my essential observations about those seven business models.

Despite significant differences in the size of their respective capital base, all the seven businesses hold most of their capital in the highly liquid financial form: cash or tradable financial securities. Their main source of financing is definitely the additional paid-in equity. Now, some readers could ask: how the hell is it possible for the additional paid-in equity to make more than the value of assets, like 193%? When a business accumulates a lot of operational losses, they have to be subtracted from the incumbent equity. Additions to equity serve as a compensation of those losses. It seems to be a routine business practice in biotech.

Now, I am going to go slightly conspiracy-theoretical. Not much, just an inch. When I see businesses such as Soligenix, where cumulative losses, and the resulting additions to equity amount to teen times the value of assets, I am suspicious. I believe in the power of science, but I also believe that facing a choice between using my equity to compensate so big a loss, on the one hand, and using it to invest into something less catastrophic financially, I will choose the latter. My point is that cases such as Soligenix smell scam. There must be some non-reported financial interests in that business. Something is going on behind the stage, there.  

In my previous update, titled ‘An odd vector in a comfortably Apple world’, I studied the cases of Tesla and Apple in order to understand better the phenomenon of outlier events in technological change. The short glance I had on those COVID-vaccine-involved biotechs gives me some more insight. Biotech companies are heavily scientific. This is scientific research shaped into a business structure. Most of the biotech business looks like an ever-lasting debut, long before breaking even. In textbooks of microeconomics and management, we can read that being able to run the business at a profit is a basic condition of calling it a business. In biotech, it is different. Biotechs are the true outliers, nascent at the very juncture of cutting-edge science, and business strictly spoken. This is how outliers emerge: there is some cool science. I mean, really cool, the one likely to change the face of the world. Those mRNA biotechnologies are likely to do so. The COVID vaccine is the first big attempt to transform those mRNA therapies from experimental ones into massively distributed and highly standardized medicine. If this stuff works on a big scale, it is a new perspective. It allows fixing people, literally, instead of just curing diseases.

Anyway, there is that cool science, and it somehow attracts large amounts of cash. Here, a little digression from the theory of finance is due. Money and other liquid financial instruments can be seen as risk-absorbing bumpers. People accumulate large monetary balances in times and places when and where they perceive a lot of manageable risk, i.e. where they perceive something likely to disrupt the incumbent business, and they want to be on the right side of the disruption.

Cultural classes

Some of my readers asked me to explain how to get in control of one’s own emotions when starting their adventure as small investors in the stock market. The purely psychological side of self-control is something I leave to people smarter than me in that respect. What I do to have more control is the Wim Hof method (https://www.wimhofmethod.com/ ) and it works. You are welcome to try. I described my experience in that matter in the update titled ‘Something even more basic’. Still, there is another thing, namely, to start with a strategy of investment clever enough to allow emotional self-control. The strongest emotion I have been experiencing on my otherwise quite successful path of investment is the fear of loss. Yes, there are occasional bubbles of greed, but they are more like childish expectations to get the biggest toy in the neighbourhood. They are bubbles, which burst quickly and inconsequentially. The fear of loss is there to stay, on the other hand.    

This is what I advise to do. I mean this is what I didn’t do at the very beginning, and fault of doing it I made some big mistakes in my decisions. Only after some time (around 2 months), I figured out the mental framework I am going to present. Start by picking up a market. I started with a dual portfolio, like 50% in the Polish stock market, and 50% in the big foreign ones, such as US, Germany, France etc. Define the industries you want to invest in, like biotech, IT, renewable energies. Whatever: pick something. Study the stock prices in those industries. Pay particular attention to the observed losses, i.e., the observed magnitude of depreciation in those stocks. Figure out the average possible loss, and the maximum one. Now, you have an idea of how much you can lose in percentage. Quantitative techniques such as mean-reversion or extrapolation of the past changes can help. You can consult my update titled ‘What is my take on these four: Bitcoin, Ethereum, Steem, and Golem?’ to see the general drift.

The next step is to accept the occurrence of losses. You need to acknowledge very openly the following: you will lose money on some of your investment positions, inevitably. This is why you build a portfolio of many investment positions. All investors lose money on parts of their portfolio. The trick is to balance losses with even greater gains. You will be experimenting, and some of those experiments will be successful, whilst others will be failures. When you learn investment, you fail a lot. The losses you incur when learning, are the cost of your learning.

My price of learning was around €600, and then I bounced back and compensated it with a large surplus. If I take those €600 and compare it to the cost of taking an investment course online, e.g. with Coursera, I think I made a good deal.

Never invest all your money in the stock market. My method is to take some 30% of my monthly income and invest it, month after month, patiently and rhythmically, by instalments. For you, it can be 10% or 50%, which depends on what exactly your personal budget looks like. Invest just the amount you feel you can afford exposing to losses. Nail down this amount honestly. My experience is that big gains in the stock market are always the outcome of many consecutive steps, with experimentation and the cumulative learning derived therefrom.

General remark: you are much calmer when you know what you’re doing. Look at the fundamental trends and factors. Look beyond stock prices. Try to understand what is happening in the real business you are buying and selling the stock of. That gives perspective and allows more rational decisions.  

That would be it, as regards investment. You are welcome to ask questions. Now, I shift my topic radically. I return to the painful and laborious process of writing my book about collective intelligence. I feel like shaking things off a bit. I feel I need a kick in the ass. The pandemic being around and little social contacts being around, I need to be the one who kicks my own ass.

I am running myself through a series of typical questions asked by a publisher. Those questions fall in two broad categories: interest for me, as compared to interest for readers. I start with the external point of view: why should anyone bother to read what I am going to write? I guess that I will have two groups of readers: social scientists on the one hand, and plain folks on the other hand. The latter might very well have a deeper insight than the former, only the former like being addressed with reverence. I know something about it: I am a scientist.

Now comes the harsh truth: I don’t know why other people should bother about my writing. Honestly. I don’t know. I have been sort of carried away and in the stream of my own blogging and research, and that question comes as alien to the line of logic I have been developing for months. I need to look at my own writing and thinking from outside, so as to adopt something like a fake observer’s perspective. I have to ask myself what is really interesting in my writing.

I think it is going to be a case of assembling a coherent whole out of sparse pieces. I guess I can enumerate, once again, the main points of interest I find in my research on collective intelligence and investigate whether at all and under what conditions the same points are likely to be interesting for other people.

Here I go. There are two, sort of primary and foundational points. For one, I started my whole research on collective intelligence when I experienced the neophyte’s fascination with Artificial Intelligence, i.e. when I discovered that some specific sequences of equations can really figure stuff out just by experimenting with themselves. I did both some review of literature, and some empirical testing of my own, and I discovered that artificial neural networks can be and are used as more advanced counterparts to classical quantitative models. In social sciences, quantitative models are about the things that human societies do. If an artificial form of intelligence can be representative for what happens in societies, I can hypothesise that said societies are forms of intelligence, too, just collective forms.

I am trying to remember what triggered in me that ‘Aha!’ moment, when I started seriously hypothesising about collective intelligence. I think it was when I was casually listening to an online lecture on AI, streamed from the Massachusetts Institute of Technology. It was about programming AI in robots, in order to make them able to learn. I remember one ‘Aha!’ sentence: ‘With a given set of empirical data supplied for training, robots become more proficient at completing some specific tasks rather than others’. At the time, I was working on an article for the journal ‘Energy’. I was struggling. I had an empirical dataset on energy efficiency in selected countries (i.e. on the average amount of real output per unit of energy consumption), combined with some other variables. After weeks and weeks of data mining, I had a gut feeling that some important meaning is hidden in that data, only I wasn’t able to put my finger precisely on it.

That MIT-coined sentence on robots triggered that crazy question in me. What if I return to the old and apparently obsolete claim of the utilitarian school in social sciences, and assume that all those societies I have empirical data about are something like one big organism, with different variables being just different measurable manifestations of its activity?

Why was that question crazy? Utilitarianism is always contentious, as it is frequently used to claim that small local injustice can be justified by bringing a greater common good for the whole society. Many scholars have advocated for that claim, and probably even more of them have advocated against. I am essentially against. Injustice is injustice, whatever greater good you bring about to justify it. Besides, being born and raised in a communist country, I am viscerally vigilant to people who wield the argument of ‘greater good’.

Yet, the fundamental assumptions of utilitarianism can be used under a different angle. Social systems are essentially collective, and energy systems in a society are just as collective. There is any point at all in talking about the energy efficiency of a society when we are talking about the entire intricate system of using energy. About 30% of the energy that we use is used in transport, and transport is from one person to another. Stands to reason, doesn’t it?

Studying my dataset as a complex manifestation of activity in a big complex organism begs for the basic question: what do organisms do, like in their daily life? They adapt, I thought. They constantly adjust to their environment. I mean, they do if they want to survive. If I settle for studying my dataset as informative about a complex social organism, what does this organism adapt to? It could be adapting to a gazillion of factors, including some invisible cosmic radiation (the visible one is called ‘sunlight’). Still, keeping in mind that sentence about robots, adaptation can be considered as actual optimization of some specific traits. In my dataset, I have a range of variables. Each variable can be hypothetically considered as informative about a task, which the collective social robot strives to excel at.

From there, it was relatively simple. At the time (some 16 months ago), I was already familiar with the logical structure of a perceptron, i.e. a very basic form of artificial neural network. I didn’t know – and I still don’t – how to program effectively the algorithm of a perceptron, but I knew how to make a perceptron in Excel. In a perceptron, I take one variable from my dataset as output, the remaining ones are instrumental as input, and I make my perceptron minimize the error on estimating the output. With that simple strategy in mind, I can make as many alternative perceptrons out of my dataset as I have variables in the latter, and it was exactly what I did with my data on energy efficiency. Out of sheer curiosity, I wanted to check how similar were the datasets transformed by the perceptron to the source empirical data. I computed Euclidean distances between the vectors of expected mean values, in all the datasets I had. I expected something foggy and pretty random, and once again, life went against my expectations. What I found was a clear pattern. The perceptron pegged on optimizing the coefficient of fixed capital assets per one domestic patent application was much more similar to the source dataset than any other transformation.

In other words, I created an intelligent computation, and I made it optimize different variables in my dataset, and it turned out that, when optimizing that specific variable, i.e. the coefficient of fixed capital assets per one domestic patent application, that computation was the most fidel representation of the real empirical data.   

This is when I started wrapping my mind around the idea that artificial neural networks can be more than just tools for optimizing quantitative models; they can be simulators of social reality. If that intuition of mine is true, societies can be studied as forms of intelligence, and, as they are, precisely, societies, we are talking about collective intelligence.

Much to my surprise, I am discovering similar a perspective in Steven Pinker’s book ‘How The Mind Works’ (W. W. Norton & Company, New York London, Copyright 1997 by Steven Pinker, ISBN 0-393-04535-8). Professor Steven Pinker uses a perceptron as a representation of human mind, and it seems to be a bloody accurate representation.

That makes me come back to the interest that readers could have in my book about collective intelligence, and I cannot help referring to still another book of another author: Nassim Nicholas Taleb’s ‘The black swan. The impact of the highly improbable’ (2010, Penguin Books, ISBN 9780812973815). Speaking from an abundant experience of quantitative assessment of risk, Nassim Taleb criticizes most quantitative models used in finance and economics as pretty much useless in making reliable predictions. Those quantitative models are good solvers, and they are good at capturing correlations, but they suck are predicting things, based on those correlations, he says.

My experience of investment in the stock market tells me that those mid-term waves of stock prices, which I so much like riding, are the product of dissonance rather than correlation. When a specific industry or a specific company suddenly starts behaving in an unexpected way, e.g. in the context of the pandemic, investors really pay attention. Correlations are boring. In the stock market, you make good money when you spot a Black Swan, not another white one. Here comes a nuance. I think that black swans happen unexpectedly from the point of view of quantitative predictions, yet they don’t come out of nowhere. There is always a process that leads to the emergence of a Black Swan. The trick is to spot it in time.

F**k, I need to focus. The interest of my book for the readers. Right. I think I can use the concept of collective intelligence as a pretext to discuss the logic of using quantitative models in social sciences in general. More specifically, I want to study the relation between correlations and orientations. I am going to use an example in order to make my point a bit more explicit, hopefully. In my preceding update, titled ‘Cool discovery’, I did my best, using my neophytic and modest skills in programming, the method of negotiation proposed in Chris Voss’s book ‘Never Split the Difference’ into a Python algorithm. Surprisingly for myself, I found two alternative ways of doing it: as a loop, on the one hand, and as a class, on the other hand. They differ greatly.

Now, I simulate a situation when all social life is a collection of negotiations between people who try to settle, over and over again, contentious issues arising from us being human and together. I assume that we are a collective intelligence of people who learn by negotiated interactions, i.e. by civilized management of conflictual issues. We form social games, and each game involves negotiations. It can be represented as a lot of these >>

… and a lot of those >>

In other words, we collectively negotiate by creating cultural classes – logical structures connecting names to facts – and inside those classes we ritualise looping behaviours.

Money being just money for the sake of it

I have been doing that research on the role of cities in our human civilization, and I remember the moment of first inspiration to go down this particular rabbit hole. It was the beginning of March, 2020, when the first epidemic lockdown has been imposed in my home country, Poland. I was cycling through streets of Krakow, my city, from home to the campus of my university. I remember being floored at how dead – yes, literally dead – the city looked. That was the moment when I started perceiving cities as something almost alive. I started wondering how will pandemic affect the mechanics of those quasi-living, urban organisms.

Here is one aspect I want to discuss: restaurants. Most restaurants in Krakow turn into takeouts. In the past, each restaurant had the catering part of the business, but it was mostly for special events, like conferences, weddings and whatnot. Catering was sort of a wholesale segment in the restaurant business, and the retail was, well, the table, the napkin, the waiter, that type of story. That retail part was supposed to be the main one. Catering was an addition to that basic business model, which entailed a few characteristic traits. When your essential business process takes place in a restaurant room with tables and guests sitting at them, the place is just as essential. The location, the size, the look, the relative accessibility: it all played a fundamental role. The rent for the place was among the most important fixed costs of a restaurant. When setting up business, one of the most important questions – and risk factors – was: “Will I be able to attract sufficiently profuse customers to this place, and to ramp up prices sufficiently high to as to pay the rent for the place and still have satisfactory profit?”. It was like a functional loop: a better place (location, look) meant more select a clientele and higher prices, which required to pay a high rent etc.

As I was travelling to other countries, and across my own country, I noticed many times that the attributes of the restaurant as physical place were partly substitute to the quality of food. I know a lot of places where the customers used to pretend that the food is excellent just because said food was so strange that it just didn’t do to say it is crappy in taste. Those people pretended they enjoy the food because the place was awesome. Awesomeness of the place, in turn, was largely based on the fact that many people enjoyed coming there, it was trendy, stylish, it was a good thing to show up there from time to time, just to show I have something to show to others. That was another loop in the business model of restaurants: the peculiar, idiosyncratic, gravitational field between places and customers.

In that business model, quite substantial expenses, i.e.  the rent, and the money spent on decorating and equipping the space for customers were essentially sunk costs. The most important financial outlays you made to make the place competitive did not translate into any capital value in your assets. The only way to do such translation was to buy the place instead of renting it. Advantageous, long-term lease was another option. In some cities, e.g. the big French ones, such as Paris, Lyon or Marseille, the market of places suitable for running restaurants, both legally and physically, used to be a special segment in the market of real estate, with its own special contracts, barriers to entry etc.   

As restaurants turn into takeouts, amidst epidemic restrictions, their business model changes. Food counts in the first place, and the place counts only to the extent of accessibility for takeout. Even if I order food from a very fancy restaurant, I pay for food, not for fanciness. When consumed at home, with the glittering reputation of the restaurant taken away from it, food suddenly tastes differently. I consume it much more with my palate and much less with my ideas of what is trendy. Preparation and delivery of food becomes the essential business process. I think it facilitates new entries into the market of gastronomy. Yes, I know, restaurants are going bankrupt, and my take on it is that places are going bankrupt, but people stay. Chefs and cooks are still there. Human capital, until recently being 50/50 important – together with the real estate aspect of the business – becomes definitely the greatest asset of the restaurants’ sector as they focus on takeout. The broadly spoken cooking skills, including the ability to purchase ingredients of good quality, become primordial. Equipping a business-scale kitchen is not really rocket science, and, what is just as important, there is a market for second-hand equipment of that kind. The equipment of a kitchen, in a takeout-oriented restaurant, is much more of an asset than the decoration of a dining room. The rent you pay, or the market price of the whole place in the real-estate market are much lower, too, as compared to classical restaurants.

What restaurant owners face amidst the pandemic is the necessity to switch quickly, and on a very short notice of 1 – 2 weeks, between their classical business model based on a classy place to receive customers, and the takeout business model, focused on the quality of food and the promptness of delivery. It is a zone of uncertainty more than a durable change, and this zone is

associated with different cash flows and different assets. That, in turn, means measurable risk. Risk in big amounts is an amount, essentially, much more than a likelihood. We talk about risk, in economics and in finance, when we are actually sure that some adverse events will happen, and we even know what is going to be the total amount of adversity to deal with; we just don’t know where exactly that adversity will hit and who exactly will have to deal with it.

There are two basic ways of responding to measurable risk: hedging and insurance. I can face risk by having some aces up my sleeve, i.e. by having some alternative assets, sort of fall-back ones, which assure me slightly softer a landing, should the s**t which I hedge against really happen. When I am at risk in my in-situ restaurant business, I can hedge towards my takeout business. With time, I can discover that I am so good at the logistics of delivery that it pays off to hedge towards a marketing platform for takeouts rather than one takeout business. There is an old saying that you shouldn’t put all your eggs in the same basket, and hedging is the perfect illustration thereof. I hedge in business by putting my resources in many different baskets.

On the other hand, I can face risk by sharing it with other people. I can make a business partnership with a few other folks. When I don’t really care who exactly those folks are, I can make a joint-stock company with tradable shares of participation in equity. I can issue derivative financial instruments pegged on the value of the assets which I perceive as risky. When I lend money to a business perceived as risky, I can demand it to be secured with tradable notes AKA bills of exchange. All that is insurance, i.e. a scheme where I give away part of my cash flow in exchange of the guarantee that other people will share with me the burden of damage, if I come to consume my risks. The type of contract designated expressis verbis as ‘insurance’ is one among many forms of insurance: I pay an insurance premium in exchange o the insurer’s guarantee to cover my damages. Restaurant owners can insure their epidemic-based risk by sharing it with someone else. With whom and against what kind of premium on risk? Good question. I can see like a shade of that. During the pandemic, marketing platforms for gastronomy, such as Uber Eats, swell like balloons. These might be the insurers of the restaurant business. They capitalize on the customer base for takeout. As a matter of fact, they can almost own that customer base.

A group of my students, all from France, as if by accident, had an interesting business concept: a platform for ordering food from specific chefs. A list of well-credentialed chefs is available on the website. Each of them recommends a few flagship recipes of theirs. The customer picks the specific chef and their specific culinary chef d’oeuvre. One more click, and the customer has that chef d’oeuvre delivered on their doorstep. Interesting development. Pas si con que ça, as the French say.     

Businesspeople have been using both hedging and insurance for centuries, to face various risks. When used systematically, those two schemes create two characteristic types of capitalistic institutions: financial markets and pooled funds. Spreading my capitalistic eggs across many baskets means that, over time, we need a way to switch quickly among baskets. Tradable financial instruments serve to that purpose, and money is probably the most liquid and versatile among them. Yet, it is the least profitable one: flexibility and adaptability is the only gain that one can derive from holding large monetary balances. No interest rate, no participation in profits of any kind, no speculative gain on the market value. Just adaptability. Sometimes, just being adaptable is enough to forego other gains. In the presence of significant need for hedging risks, businesses hold abnormally large amounts of cash money.

When people insure a lot – and we keep in mind the general meaning of insurance as described above – they tend to create large pooled funds of liquid financial assets, which stand at the ready to repair any breach in the hull of the market. Once again, we return to money and financial markets. Whilst abundant use of hedging as strategy for facing risk leads to hoarding money at the individual level, systematic application of insurance-type contracts favours pooling funds in joint ventures. Hedging and insurance sort of balance each other.

Those pieces of the puzzle sort of fall together into a pattern. As I have been doing my investment in the stock market, all over 2020, financial markets seems to be puffy with liquid capital, and that capital seems to be avid of some productive application. It is as if money itself was saying: ‘C’mon, guys. I know I’m liquid, and I can compensate risk, but I am more than that. Me being liquid and versatile makes me easily convertible into productive assets, so please, start converting. I’m bored with being just me, I mean with money being just money for the sake of it’.