Gold with the wind
March 28th, 2012 by Claude Crampes, Toulouse School of EconomicsThe megawatts invested in capacity to produce electricity from renewable energy and the megawatts/hours produced from this capacity are growing faster than expected within the European Union. That is definitely good news for climate change mitigation. However, what about the cost of such policy? Evaluating the cost of renewable sources within the energy mix is not an easy task. A key ingredient that needs to be considered is the intermittency of renewable energy sources. It affects the financial balancing of the electricity industry where fossil fuel plants are to be built and operated to replace intermittent sources at dates they are not producing.
The point is that among the highly heterogeneous renewable energy sources (hereafter, RES) the two most subsidized sources are wind and solar energy. They produce intermittently, partly in a cyclical way, partly in a random way. In practical terms, it means that whatever large the installed capacity and total production of RES, there is no guarantee that power will be available at the time we need it. At some dates, production will be larger than demand, at other dates it will be smaller. One could ask market authorities to solve the adjustment problem by allowing energy prices to move up and down at high speed. This actually cannot work as long as consumers do not receive the price signal. And even if they could know the spot price thanks to some electronic beeper, they still would have to adapt their energy withdrawal in real time, which requires specific electronic pieces of equipment. Absent all these smart helpers, the solutions that could be envisaged are technical. They combine i) energy transfers, either from one location to another (transmission) or from one date to another (storage), ii) demand adaptation (black-outs, brown-outs, load-shedding), and iii) additional production from non-intermittent sources. Let us now focus on the investment in fossil fuel plants as a back-up for the intermittency of solar and wind energy.
Consider the near future where one can expect that the RES investment costs will be low enough to compete against the thermal technologies. In this framework, both types are then installed, the former because of their close-to-zero operating cost, the latter because RES are intermittent. In a world of perfect information of consumers on the state of nature at production nodes, electricity should then have a low price when RES are available and a high price when they must be replaced by plants burning a costly fuel (including CO2 certificates). The optimal mix of technologies would combine RES and fossil-fuel plants on the basis of these state-contingent prices and both types of producers would balance their accounts. Actually, as long as the cost of installing smart appliances at consumption nodes coupled with devices from the Information and Communication Technologies all along the grid is too high compared to the welfare increase from introducing state-contingent prices, consumers continue to be offered only one price whether the intermittent sources are producing or not. They consequently consume too little energy when the intermittent sources are available and too much when they are not. Suppose that the equilibrium (or the regulated) non-contingent retail price is an average value of the marginal costs of production in the different types of generation plants. Then the retail price is above the RES cost and below the thermal plant cost. It results that the generators using fossil sources of energy will lose money so that they prefer to leave the industry.
Anticipating the future where RES will be able to compete against fossil fuel plants without subsidies, we therefore have to consider structural and legal solutions to implement the optimal energy mix constrained by non-contingent prices. Under free entry and exit, in order to keep generators using fossil fuel in the market, the price should not be lower than the long run marginal cost of their MWh. It is as if consumers will have to pay for a warranty of electricity supply. The drawback of this solution is that the owners of plants using intermittent energy pocket a profit equal to the difference between the long run marginal cost of electricity from fossil fuel and the long run marginal cost of electricity from intermittent energy. Consumers pay for being insured against random supply and the money they pay accrues to those who create randomness. With this profit in view, RES producers invest even more, which increases the intermittency of the installed mix.
Two basic public policies can reduce the rent assigned to intermittent energy producers and restore the second best energy mix. The first one consists of taxing RES and subsidizing thermal plants in order to balance the budget of each type of technology, the reverse of today policy. The second one consists of a mandatory technological mix in order to produce a non-random energy outflow. Each producer should either control the two technologies or buy an insurance contract that secures energy supply any time. These legal arrangements would force both the incumbents and new entrants to guarantee energy provision whatever the state of nature at the production plant locations. Both policies have their drawbacks. The first one, more market based oriented, comes at the cost of levying and redistributing public funds. The second one, more in line with command-and-control regulation, restricts the firms’ flexibility in their technology choice.
An intermediary solution is to reinforce the role of suppliers in providing final consumers with both energy and insurance through a menu of contracts. Yet this is for a more distant future because the provision of some consumers with cheap intermittent electricity and others with expensive reliable electricity within the same distribution network will necessitate more than smart meters. Consumption appliances will have to be equipped with microprocessors and connected through the “internet of things” to the suppliers. At periods of scarce energy, the suppliers will then be able to (partly) disconnect Mr. A who did not sign a “non-random contract” while still fully supplying Ms. B, his nearest neighbor, who did it.
Stefan Ambec, Toulouse School of Economics (INRA-LERNA-IDEI), and University of Gothenburg, and Claude Crampes, Toulouse School of Economics (GREMAQ-IDEI)
P.S. The authors have built an analytical economic modeling of the first-best and second-best production mix when intermittent sources can be combined with reliable sources. See “Electricity provision with intermittent sources of energy” published in Resource and Energy Economics 34 (2012) 319– 336 at http://www.sciencedirect.com/science/article/pii/S0928765512000024
March 28th, 2012 at 8:42 pm
Start with a theoretical solution that assumes away certain difficulties, and then add in the difficulties and find practical modifications of that solution. Perhaps that is the approach taken above. But, I don’t find the starting point to be clear. So consider the following.
1.Assume players are not risk averse and enough consumers are price elastic so that the spot price clears the market below VOLL. Suppose we tax CO2 emissions optimally.
Market solution (1): Set the spot price to the marginal cost of generation, or when supply is exhausted, to the marginal value of elastic consumers (who respond to price). This will induce the optimal mix of generation technologies. I.e. the market works, except for needing a CO2 tax.
2. Assume as before, but price sometimes exceed VOLL.
Semi-market solution (2): The regulator should cap the price at VOLL and shed load at that point. Again we get the best possible mix of technology given the realities of consumer demand.
3. Unfortunately such markets produce extreme spiky, and hence very risky, prices – generators may make a lot of money every 10 or so years, and lose money in between. Also such market give generators enormous market power whenever prices are close to spiking.
Solution (3): The above plus, require generation to sell (load following) price hedges with a strike price of, say $300/MWh, to collective consumers.
Discussion: Hedges do not interfere with incentives. There will still be price spikes to VoLL, but these will not cause market-risk (only performance risk), and they will not cause market power.
4. Two final problems: (1) The selected VoLL (even if optimal—no chance of that) may result in a levels of installed capacity and reliability that engineers and politicians find unacceptable. This will cause a regulatory crisis. (2) Investment coordination is still quite poor, and hence inefficient.
Solution (4): The above plus implement a capacity market with a politically (and engineeringly) acceptable capacity target.
Discussion: This solves the final too problems. But capacity targets, like VoLL, are administratively determined and hence imperfect. There is an strong suspicion that capacity targets are less accurate than VoLL guesses and that they are too high. Suppose so. Suppose 5% extra capacity is targeted. This is the cheapest form of capacity (since it is almost never used) and it requires essentially no fuel and no extra wires or consumer billing costs. The resulting increase in the retail price is about ½ of 1%. This is negligible compared to the wastefulness of the current system of huge and variable subsidies.
April 11th, 2012 at 9:25 pm
The first two solutions offered in this article to avoid a loss of competitiveness of thermal power plants seem to be problematic. Subsidizing thermal power plants instead of renewable energy would be extremely difficult. As a result of intermittence, renewable energies would not produce the same amount of electricity in a given period of time. That would lead to thermal power plants not producing and selling the same amount of electricity. The subsidies that should compensate for the losses would therefore have to be adjusted continuously to reach a balance.
The second solution, consisting in forcing energy producers to use a mix in order to always have a given power available, raises other complication. It is not always easy for producers to diversify their means of production because that requires them to master new technologies. This leads to new cost in research and development and they do not have the same efficiency as other producers because they lack experience in those new technologies. Another close solution would be to force producers of renewable energy to unite through contract with thermal energy producers and pay them. This is in fact very close to the third solution of the article.
This solution is probably the best one. If the suppliers offers insurance to their clients, the cost of these insurances would allow suppliers to pay for thermal power plants to be ready to produce at any time. The article raises the problem of needing extremely accurate consumption appliances, in order to disconnect people who did not pay for the insurance but not the people who did. This considers that some people will probably choose not to take these insurances and be disconnected in period of scarce energy. But there is another solution. If energy suppliers only offer contracts with these insurance included, there will be no need for these new appliances and it will even lower the cost of the insurance for everyone. Of course, it raises the problem of people not being ready to be forced to pay for things they do not want. But, as social security forces you to pay for it, even if you are not sick, it is not inconceivable to imagine a regulation forcing people to pay for energy production at times they do not use it.
April 12th, 2012 at 11:01 pm
On the contrary of what has been said, I think that a consumer is totally risk-
averse. He would like to have throughout all the year and at the same price power
regardless to the hazard implied. Therefore, it requires that the price of electricity
has to be almost constant throughout the year for consumers: the hazard implied by
the weather should not be charged to them. Thus the increasing or decreasing of the
electricity price should only be seen at long term and not all year round. Only, it’s
impossible to have a constant price for the following reasons:
Firstly, for external causes: inflation for example. Thus we can see at long term an
increasing/decreasing of the electricity price even if it will become quasi-constant (with
a poor volatility).
Secondly, because electricity market is changing with the augmentation of the
numbers of customers for example. Thus it follows the rule offer/demand.
I think all consumers should purchase a « non-random contract » as said,
meaning that there should be only one type of contract for all consumers. It’s almost a
matter of public order even if that means that the price goes up slightly.
Indeed, I think that it’s impossible to accept the fact that at periods of scarce
energy, “Mr. A”, not protected against the randomness, will not have electricity like his
neighbour. Everybody will understand this same problem if we refuse to Mr. A to buy
bread because he didn’t subscribe to a non-random contract over the randomness of the
wheat production: this kind of contract doesn’t exist for other first needs. Why does it
exist for electricity?
However, with the looks worn by the current politicians and the global financial
markets, it would be good to ensure a full transparency on the volatility of electricity,
which has now become just a financial asset (see commodities). It should also be good
to inform consumers about the sources of the electricity generally, explaining whether it
comes from renewable energy or power plant, etc. A consumer should want to know if
it is a scarce period or not for electricity (even if at the end he will not know how much
electricity from power plants he will have used because of the non-traceability of the
electricity) Also, it’s important to inform the consumer on the randomness of energy
production.
But it’s important to assume the risk. I think it should totally be assumed by the
state or the insurance of the electricity providers. I do not necessarily approve this kind
of « intermediate solution ». Instead, a mixture of both solutions could be interesting.
Subsidizing and taxing RES thermal plants and at the same time use a mandatory
technology mix in order to produce a non-random energy outflow.
On the other hand, on a non-consumer perspective, but rather from the buyer’s
perspective of electricity trade by traders on commodities, the interest of a non-random
energy solutions totally different. The volatility of the electricity price will then certainly
decrease.
Last but not least, should we not consider this kind of subject as a policy matter
for the EU. In our days, there is far too much gap between the choices made in the
different countries of the EU concerning the sources of electricity. Germany and France,
for example, who show a close and friendly policy, tend to differ on this subject.
Even though I may have a utopian spirit, I think that a European directive should impose
a constant price of electricity to all consumers. Indeed, as I said before, in my opinion, it
is a question of first needs, which can’t depend on the hazard and can’t be different from
a customer to an other.
Benjamin BENOUDIS
Student at Mines Paristech, Engineering school
Quantitative finance option
July 1st, 2012 at 3:44 am
The point about energy generated from wind and solar cannot be guranteed is a good one. We cannot rely on the sun or the wind, we must always have an alternative energy source that is not subject to the vagaries of the weather. There must be sufficient capacity to provide electricity during dull, overcast, windless days and nights. In the UK the government has realised that the backbone of electricity production has to be somehing we can rely on 365 days and nights a year and they have decided that the only solution is to increase the reliance on nuclear power.
As well as wind and solar, there will also be an investment in electricity generated from the tides. This is all well and good, but until the UK has increased its nuclear capacity we run an ever increasing risk of power outages. Unfortunately it takes many years to get these power plants up and running, and one questions whether the Government has left this all a little to late. Time will tell.