VTT-R-04621-16 Electricity Market Designs and Flexibility (2016)

Electricity market designs and flexibility. VTT Technical Research Centre of Finland, Research Report VTT-R-04621-16, Espoo, 31 October 2016. Authors: Juha Forsström, Göran Koreneff, Lassi Similä. Produced under the FLEXe programme (Tekes). 50 pages (per the report’s cover sheet), including an appendix Q&A.

Full title: Electricity market designs and flexibility

Report code: VTT-R-04621-16

Programme: FLEXe (Flexible Energy Systems), funded by Tekes (the Finnish Funding Agency for Innovation)

Summary

The report asks how electricity markets can be organised so that flexibility is efficiently tackled, treating market design rather than technology as the subject. It concludes that current advanced market designs already allow some trading of flexible resources, that many identified shortcomings stem from general characteristics of electricity markets, and that explicit conclusions on an adequate design are therefore hard to draw. It discusses three directions: capacity mechanisms, more flexible components in existing market systems, and fundamentally new market solutions. Most attention goes to the two-market system of Keay (2016) and its applicability to the Nordic market. The perspective is a 2016 Finnish/Nordic one.

The report is structured in five chapters plus an appendix: (1) introduction, (2) shortcomings in the current market design and suggested solutions, (3) essays on the Nordic/Finnish market (capacity mechanisms, low prices, loss of condensing power), (4) the two-market design with a VTT simulation, and (5) a thought experiment for a separate renewables market in the NordPool environment with a Wind System Operator. The Appendix Q&A tabulates the claims, arguments, counter-arguments and suggested solutions.

Four claims about market shortcomings

The report was prepared in internal author meetings that discussed claims often made about electricity markets. The stated goal was not to give direct answers but to open up the field of research. Four claims are set out, each with arguments and counter-arguments:

1. “Market prices are too low.” Subsidised low-variable-cost capacity pushes prices below short-run marginal cost, causing losses and capacity exit, a vicious circle in which only subsidised capacity survives. The counter-argument is that low prices are normal and signal overcapacity. The report links this to the “missing money” problem of energy-only markets.
2. “Market prices are too high.” Market power in oligopolistic markets, extreme in system-stress hours because inflexible demand cannot respond to high prices; generators are said to enjoy windfall profits. The counter-argument is that high prices signal capacity deficit and that profits reflect right strategic choices.
3. “Markets do not produce enough flexibility.” Much power cannot adjust production in the short term, leaving flexibility to system operator functions; in the extreme this could endanger system security. The counter-argument is that the market produces enough flexibility, or that flexibility should be left to the system operator with tailor-made mechanisms.
4. “Markets do not produce enough capacity.” Attributed to oligopolistic behaviour and entry barriers. The counter-argument is that markets lead to the optimal amount of capacity, that inertia in capacity expansion is normal, and that the problem is mainly caused by feed-in tariffs dampening prices.

The authors note that one cannot prove a price outcome “wrong” (price movements always create winners and losers) and that a low-price period cannot easily be separated into “fundamental flaw” versus normal overcapacity, so one can usually find arguments both for and against reform. The report gives Nord Pool monthly prices falling since 2011, with a low of 9.55 EUR/MWh in July 2015 (Figure 3), and says the claim “too low” has dominated the Nordic discussion recently.

Suggested solutions (Chapter 2)

Table 2 lists capacity mechanisms; more flexible components in market systems (15-minute or shorter markets; a pool with centralised optimisation of flexibility resources); the “two-market design”; and “platform markets”. Capacity mechanisms are explained by three electricity-specific characteristics: lack of price-responsive demand, limits on storing the product, and vulnerability to market power in high-demand periods. The report notes these consequences may have worsened with massive subsidised RES take-up in Europe, and suggests that relying only on traditional capacity mechanisms may not be adequate for flexibility.

• Capacity mechanism types listed: capacity payments; strategic reserves (experience in Sweden and Finland); capacity requirements (US, typically about a 15 % reserve margin above estimated peak); capacity markets; reliability options.
• UK capacity market (first run late 2014): used as a case study. In the first T-4 auction (winter 2018/19) 49.26 GW (derated) was procured, of which 64 % existing capacity, 14 % refurbishments, 16 % pre-refurbishment, about 5 % new build and 1 % demand-side resources (National Grid 2015). The authors read the demand-side outcome as discouraging.
• Short-term markets: citing IEA (2016), improvements suggested are high geographical resolution to price congestion, uniform prices for real-time balancing energy, and continuous updating in the last hours before operation. The report also stresses that temporal resolution (hourly, 15-minute, 5-minute) is set by market rules rather than by a law of nature.
• Scarcity pricing and reliability: capping price at the value of lost load minimises the social cost of market failure but VoLL is hard to value and can reach dozens of thousands of euros (IEA 2016). Typical reliability standards are 10-15 % above peak-load capacity. IEA (2016) recommends higher price caps, ex ante market power mitigation and regulation of scarcity price formation, and concludes some form of capacity mechanism may be needed as a safety net during the transition.
• Fingrid 2016 development actions (Finnish TSO, following a spring 2016 consultation): pilot of a 30-minute intraday gate closure, publication of real-time balancing prices in scarcity situations from November 2016, two pilots on third-party reserve-market participation, and 5 MW bid size where electronic subscription is in place.

Nordic essays (Chapter 3)

Chapter 3 is an essay-style discussion from Nordic/Finnish stakeholder perspectives.

• The Nordic market is energy-only, but Finland and Sweden use strategic reserves, and renewable support schemes (green certificates in Norway and Sweden, price guarantees in Denmark and Finland) are said to work similarly to capacity mechanisms. Missing money “is actually solved if demand response functions as a price setter.”
• Condensing (regulation-capable) power plants are threatened by low prices and low need for them; several have closed. Calculations by Koreneff et al. (2015) indicate Nordic condensing demand might fall to 4-6 TWh by 2035.
• The strategic reserve in Finland/Sweden has fallen from 600 MW to about 300 MW in ten years without problems; Finland’s record peak load of 15,100 MW (7 January 2016) did not require reserves. The Nordic spot price ceiling at the time was 3000 EUR/MWh.
• The authors judge low prices “not a problem per se” and see stale, stiff end-user tariffs as a bigger issue for flexibility preparedness. They favour extending strategic-reserve studies to 2025 and 2030, say capacity mechanisms “might become necessary in the future”, and call demand response among the most cost-effective tools, urging removal of regulatory and market barriers.

The Keay (2016) two-market design (Chapter 4)

The report analyses the two-market design proposed by Malcolm Keay (OIES paper EL17, 2016), presented as intended to upgrade the UK market. A new design must incentivise customer flexibility, accommodate zero-marginal-cost technologies, have a vision for ending economic support of them, incentivise timely flexible-capacity investment, and give sustainable price levels. The report frames zero-marginal-cost generation as a “fixed price world” in which the market is effectively a capacity, not energy, market (Helm 2015).

Firm market

• Product: firm power available on demand; retail price higher than “as available” electricity
• Participants: like the original spot market, open only to generators with flexible generation technology; VRE generators can enter by contracting with a flexible generator to form a hybrid firm and flexible unit
• Pricing must cover all generation costs; additional capacity mechanisms can be used if capacity does not reach the needed level

Non-firm market

• Product: energy available when circumstances allow generation
• Participants: open to all generators, firm and variable, but intended in the first place for zero-marginal-cost technology (wind, solar PV)
• Price: while support is given, set by a regulator on the basis of expected long-term marginal cost of capacity; the long-run aim is direct sales without support at a market price

Retail and interaction

• In Keay’s basic retail model, a consumer holds both a firm and a non-firm contract sized to maximum demand, behind one meter. The consumer pays the non-firm price so long as non-firm generation exceeds the demand of that class, and a share of the two prices otherwise (half non-firm generation means half higher, half lower price). An hourly meter is a prerequisite.
• The report suggests information on availability and automatic load control could be a prerequisite for a non-firm contract; a customer with two meters could put only flexible loads behind the non-firm meter.
• One network and one system operator: non-firm power is dispatched together with firm power, automatically except where curtailment is applied.

Why the authors see it as a demand-side incentive

The authors say what is missing today is the incentive to turn the demand side’s potential flexibility into an operable state. A lower-priced but restricted-availability product is meant to incentivise flexibility “on an everyday basis”, making demand-side flexibility investments viable. The report gives no analysis of peak/off-peak price spreads under the present single market.

VTT simulation (Section 4.5)

A simplified model uses a representative day of 24 one-hour steps and stylised plants (Table 3, not real plants) plus randomly generated synthetic wind variability. Sustainability is judged by comparing levelised unit fixed costs against short-term unit net revenues per plant.

• Spot market without VRE: works as designed; only the marginal oil-fired gas turbines show negative economics, which the authors attribute to the representative demand curve not revealing rare scarcity hours.
• Spot market with VRE: VRE lowers prices and plant economics deteriorate. “Only nuclear plants show positive result”; not even the VRE technologies themselves are sustainable. The authors add that the comparison is not quite fair because the firm fleet structure is unchanged and would adapt over time.
• Two-market, static customer flexibility: consumer flexibility modelled as the same demand decrease in every hour. Non-firm energy stays at the lowest price in all but one hour. Firm-market prices are only slightly different from the spot-with-VRE prices; the average firm supply is lower and capacity was not adapted. After capacity adaptation the authors expect firm-market economics to equal the original spot market.
• Two-market, refined flexibility: flexibility concentrated in the middle of the day (peak) makes residual demand look like variations around the daily average rather than a two-peak curve, an easier profile for the firm market to follow. But the model cannot price dynamic features, and the two designs are “almost similar in economic respect”; only nuclear appears sustainable in this case.
• Caveat: the model calculates energy costs statically with no dynamic control restrictions, start/stop costs or ramping costs, and is “currently too simple” to show the design’s features rigorously. The authors call for more work and describe the starting point as intriguing.

The Wind System Operator (WSO) thought experiment (Chapter 5)

Chapter 5 is a thought experiment on how a separate intermittent (wind) market could work using mainly today’s Finnish and Nordic tools, splitting the markets with a firewall. Due to shortened project resources, no model calculations were made on feasibility or money flows.

• WSO: holds system balance responsibility for the wind generators and for end-users using the wind power. Price tariff is constant for load that follows generation and high for load that does not; price is lower than the normal market and set by the WSO at least initially (state-guaranteed producer price differences are expected to be paid anyway); auctions might come later.
• End-user mechanics: the user receives a pro-rata share of the variable production and pays imbalances separately, buying down- or up-regulation at the wind balance market price. A user who flexes to match pays only the WSO base price. Illustrative Table 4: 5 units contracted from the normal market at 30 and 3 from the renewables market at 20, with up-regulation at 40 and down-regulation at 10. Using 9 units gives 250 in total (average 27.8); using 7 gives 200 (average 28.6).
• Producers and spot market: wind producers are responsible for their own forecast imbalances. The WSO sells surplus and buys shortfall in the normal spot market after spot prices are settled; extra spot revenues could go to replaced production and a marginal capacity fixed-cost fund. Wind producers receiving support would initially be forced into the WSO system.
• Balance settlement (four cases): WSO overproducing with down-regulation receives the down-regulation price; overproducing with up-regulation receives the lower of the WSO and normal spot price; underproducing with down-regulation pays the higher of the two spot prices; underproducing with up-regulation pays the up-regulation price. Extra balancing revenue to the system operator could fund capacity repayment or reduce balancing fees on the normal side.
• Claimed advantages: no major renovation of metering systems and minimal impact on software and data management; a more stable spot price; sharper allocation of balancing costs to those causing them. Users may also keep a fixed part of their delivery on the normal side.
• Authors’ caveats and afterthought: not all money flows or alternatives are set out; with low future wind and solar LCOE the system “might work even without subsidies”.

Key claims and caveats

• The report explicitly does not give direct answers; it maps arguments and counter-arguments and says explicit conclusions on adequate market design are challenging.
• The two-market simulation is acknowledged as too simple to demonstrate the design rigorously, and its economics results for the two-market cases are close to the spot-with-VRE case.
• The WSO proposal is a thought experiment with no financial modelling.
• The appendix flags FLEXe-related questions: whether solutions consider the quality of capacity and new technology, demand-side options, and small-scale players and prosumers.

Relevance to wiki

• Demand Response - two-market design as an everyday demand-side flexibility incentive
• Balancing Markets - missing money problem, Nordic strategic reserves, scarcity pricing
• Flexibility Market - non-firm market concept as a market-design idea for flexibility