1

ENERGY ECONOMICS

10 -25 May 2016

WHOLESALE ELECTRICITY MARKETS – part 3

MARKET INTEGRATION

PAYING FOR CAPACITY: THE ECONOMIC RATIONATE

CAPACITY REMUNERATION MECHANISM (CRM)

2

Summary MARKET INTEGRATION .............................................................................................................................. 3

Directive 2009/72/EC ............................................................................................................................... 4

Regulation 714/2009 ................................................................................................................................ 4

Framework Guidelines (FG) on CACM for electricity ................................................................................ 4

Target model ........................................................................................................................................... 4

Explicit auctions .......................................................................................................................................5

Implicit auctions .......................................................................................................................................5

Coupling projects .....................................................................................................................................5

Nordic Power Market (Nord Pool) .........................................................................................................5

MIBEL (2007) ........................................................................................................................................5

Trilateral Market Coupling (TLC) 2006--‐2010 ..................................................................................... 6

Central West European (CWE) Market Coupling .................................................................................. 6

Price Coupling of Regions (PCR) .......................................................................................................... 6

PAYING FOR CAPACITY: THE ECONOMIC RATIONATE ............................................................................. 7

Example ................................................................................................................................................. 10

CAPACITY REMUNERATION MECHANISM (CRM) .................................................................................... 13

MOTIVATION ......................................................................................................................................... 13

CLASSIFICATION ................................................................................................................................... 13

STRATEGIC RESERVE ........................................................................................................................ 13

CAPACITY OBLIGATION .................................................................................................................... 14

CAPACITY AUCTION .......................................................................................................................... 14

RELIABILITY OPTIONS ....................................................................................................................... 14

CAPACITY PAYMENTS ....................................................................................................................... 15

ACTUAL SITUATION IN EUROPE ........................................................................................................... 15

3

MARKET INTEGRATION In 1995, only two countries of Europe used to operate Power Exchanges: England & Wales and Norway.

The situation improved quite rapidly in the first 2000s until including almost all the countries in 2012:

One of the main objectives of the European Union concerning energy is the creation of an internal market.

The aim of the internal market is to create a single market without barriers to trade.

However, it has found some obstacles:

Interconnection capacity is not sufficient to support exchanges that would be economically justified

Lack of rules for the efficient use of the interconnection capacity already available

Therefore, the main issue is the lack of international agreements to trade electricity through different

countries.

The European Union has moved many steps towards this ambitious goal, first of all reinforcing the legal

framework in this sense.

4

DIRECTIVE 2009/72/EC Transmission System Operators (TSOs) cooperate through a specific European organization

(ENTSO-E).

National authorities coordinated by ACER (Agency for the Cooperation of Energy Regulators).

REGULATION 714/2009 The task of managing cross-border trade is given to ENTSO-E

o creation of Network codes to regulate the allocation of interconnection capacity and to deal

properly with congestion (Capacity Allocation and Congestion Management, CACM).

TSOs shall promote operational arrangements in order to ensure the optimum management of the

network and shall promote the development of energy exchanges, the coordinated allocation of

cross border capacity through non-discriminatory market-based solutions, paying due attention to

the specific merits of implicit auctions for short‐term allocations, and the integration of balancing

and reserve power mechanisms.

FRAMEWORK GUIDELINES (FG) ON CACM FOR ELECTRICITY Capacity allocation in the day-ahead market on the basis of an implicit auction implemented

through a single market coupling algorithm

This algorithm has to determine prices and volumes in all areas based on a marginal price principle

The implementation will take into account the role of Power Exchanges and provide for the

harmonization of gate closures for the acceptance of bids in the day ahead market

The algorithm calculates a single price for each area and for each hour (for sellers and buyers)

In case of congestion, the cost to use the interconnection lines must be equal to the difference

between the zonal prices recorded on day ahead markets of the different areas

Similar rules apply to the intraday markets for which there is also trading on a continuous basis

TARGET MODEL TSO are jointly responsible for the calculation of transmission capacity.

PX are singularly responsible for

gathering of bids, defining the

order books and for the

clearing of payments

PX are jointly responsible for

o Matching orders, using a

single algorithm for the

determination of prices

in each market area

o determining cross-

border trade

o distributing congestion rents

5

EXPLICIT AUCTIONS The Joint Allocation Office (JAO) is a joint service company of twenty Transmission System

Operators (TSOs) from seventeen countries.

The JAO constitutes a merger of CAO Central Allocation Office GmbH located in Germany and

CASC.EU S.A. (Capacity Allocation Service Company) located in Luxembourg

It mainly performs the yearly, monthly and daily auctions of transmission rights on 27 borders in

Europe and act as a fallback for the European Market Coupling.

IMPLICIT AUCTIONS Implicit auctions

Lower transaction costs (a single auction for energy and transmission rights)

Capacity allocated in function of the price differential (no inefficient flows)

Proper signal for new investments

COUPLING PROJECTS NORDIC POWER MARKET (NORD POOL)

Norwegian Power Exchange (1995) includes progressively Sweden (1996), Finland (1998), Denmark (2000),

Estonia (2010), Lithuania (2012). There is only one PX and relations between PX and TSO are regulated by

service contracts

MIBEL (2007)

It involves Spain (OMEL) and Portugal (OMIP). DAM algorithm and Intraday entrusted to OMEL and

management of derivatives assigned to OMIP. It is a regulated model.

Trilateral

6

TRILATERAL MARKET COUPLING (TLC) 2006--‐2010

Powernext in France (2001), APX in the Netherlands (1999) and Belpex in Belgium (2006). The algorithm is

assigned to a PX (Netherlands) and the relations between PX and TSO are regulated by service contracts.

CENTRAL WEST EUROPEAN (CWE) MARKET COUPLING

It involves France, Netherlands, Belgium, Germany and Luxembourg. The PX links: APX --‐ ENDEX, Belpex,

EPEX SPOT. The algorithm has been specially created and then entrusted to a PX (EPEX SPOT). Moreover

relations between PX and TSO are regulated by service contracts.

PRICE COUPLING OF REGIONS (PCR)

Initiative of seven Power Exchanges: APX--‐ENDEX, Belpex, EPEX SPOT, GME, Nord Pool Spot, OMIE and

OTE. It covers the electricity markets in Austria, Belgium, Czech Republic, Denmark, Estonia, Finland,

France, Germany, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden,

Switzerland and UK.

The initiative started in 2009 and the PCR parties signed the PCR Cooperation Agreement and PCR Co--‐

ownership Agreement in June 2012. On 4 February 2014, Price Coupling in North Western Europe was

launched: it was the first project to use the PCR solution.

It aims to develop a single price coupling solution to be used to calculate electricity prices across Europe,

and allocate cross-border capacity on a day-ahead basis.

PCR is based on three main principles: a single algorithm, robust operation and individual power exchange

accountability

1. The common algorithm will give a fair and transparent determination of day-ahead electricity

prices across Europe and allocate cross‐border capacity

2. The PCR process is based on a decentralized sharing of data, providing a robust and resilient

operation

3. The PCR Matcher and Broker service enables exchange of anonymized orders and area-to‐area

transmission capacities among the power exchanges to calculate area prices and area-to‐area

transmission flows for all included bidding areas

7

PAYING FOR CAPACITY:

THE ECONOMIC RATIONATE To explain how electricity prices vary over time, we must consider the costs involved.

Start with panel (1) of the following figure. The two straight lines show the total cost (per MW) of operating

two types of power station for different numbers of hours over a year. The vertical intercept represents the

station’s fixed costs, while the slope gives the variable cost per megawatt-hour (MWh) generated. The

peaking plant (which might represent an open-cycle gas turbine or an old plant nearing retirement) has

lower fixed costs but a high variable cost. The baseload plant

(combined-cycle gas turbines (CCGTs) are the investment of

choice in most countries where gas is available) has higher fixed

costs but a lower variable cost. If the plant is needed for more

than T hours a year, it is cheaper to build a base-load plant,

whereas the peaking plant is cheaper for less intensive use. The

thicker line segments show the lower envelope of the cost

function for this efficient plant mix.

How much capacity of each type is needed? Move down to panel (2),

which shows the load–duration curve. The hours of the year are ranked

in decreasing order of demand, and there are T hours in which the

demand is at a level of B GW or above. If the industry has B GW of base-

load plant, they will all be used for at least T hours a year. If there were

any more base-load plants, they would be used for less than T hours, and

it would have been cheaper to use peaking plants instead.

What about the total capacity? At this point, we need to remember that

the demand for electricity, and hence the load–duration curve, depends upon its price. The load–duration

curve in panel (2) is drawn on the assumption that the price of electricity is the variable cost of the marginal

plant in operation. At a price equal to the variable cost of a peaking plant, the maximum demand for

electricity would be D. Peaking plants would not be able to recover their fixed costs, however, if the price of

electricity never rose above their variable costs.

In an ideal electricity market, there would be a smooth demand curve

for power at each point in time: some consumers have to pay the

market price and would be willing to reduce their demand as the price

rose. The highest of these demand curves is shown in panel (4).The

heavy line in panel (4) is the industry’s marginal cost curve. There are B

GW of base-load plant with a variable cost of CB, and (K – B) GW of

peaking plant with a variable cost of CP. The marginal cost curve

becomes vertical at K, the industry’s total capacity. If the price were

equal to CP, the maximum demand would be equal to D GW, but since

this much cannot be generated, the price must rise to PR in order to ration demand to capacity.

8

This leads us to the price–duration curve shown in the final panel, (5). The

maximum price is PR, and the price remains above CP for the first P hours

of the year. At hour P, panel (2) shows that the demand at a price of CP is

just equal to K, there is no need to ration demand to capacity, and the

price is equal to the variable cost of the peaking plants. After T hours,

demand is low enough to be met by the baseload plants alone, and the

price drops to CB.

The shaded area just under the top of the load–duration curve (2) represents the electricity that would be

demanded at a price equal to variable cost but is not generated. In this way it flattens the peak demand to

the level of available capacity, K. Note that in a system where rationing by price was not possible, but the

available capacity was still equal to K, the shaded area would represent non-price rationing, in the form of

‘unserved energy’ (i.e. power cuts).

Finally, we can return to the top panel (1) to consider the question of cost recovery. The curved line in the

bottom left of the panel shows the rate at which plants earn revenue in the highest-demand hours of the

year—the line has a steep slope because prices are high. By hour P, this total revenue line has met the thick

line giving the total costs of a peaking plant. Since all the peaking plants are in use for at least the first P

hours of the year, this allows all of them to cover their fixed costs. Since the marginal revenue for each

additional hour’s generation up to T is equal to the variable costs of the peaking plants, the total revenue

line is superimposed on the total cost line for those plants. After T, the price drops to the variable cost of the

base-load plants, and so the slope of the total revenue line falls. By the definition of T, however, this is the

point where the two types of plants have equal total costs, and so the base-load plants will also have total

costs equal to their total revenues.

This analysis has been simplified shows that if the industry has the right amount and mix of capacity, all

plant types can recover their costs from market-based pricing. The key requirement is that peak prices must

rise above the variable costs of peaking plant. If we have too little plant, prices will be above this level for

longer, and so all types of power station will be paid more than their costs, signaling that entry will be

profitable. If there is too much capacity, prices will only exceed variable costs by a small amount, and

stations will lose money, encouraging exit.

Similar conclusions can be drawn about the mix of capacity—if there are too many peaking plants relative

to base-load capacity, prices will be high for a greater proportion of the year. This raises the earnings of

base-load stations above their costs, sending a signal that more are needed.

We should also discuss the pricing of electricity transmission. Ignoring losses for a moment, if there were no

transmission constraints, the marginal cost of power could be the same at every point in the system. If there

is a binding limit on the flow along some line, however, some more expensive output from the importing

side of the constraint must replace some cheaper output from the exporting side. If the price in each region

is equal to the marginal cost in that region, then the difference between the two prices represents the

marginal cost of the transmission constraint, and the economically correct charge for sending electricity

between the two regions.

In a meshed network, however, in which there are many possible routes between any two points, it is no

longer correct to think of just two prices. If a node on the network is ‘close’ to the constraint—the resistance

9

on the lines between that node and the constraint is relatively low—then Kirchhoff’s laws imply that a

relatively high proportion of any change in generation at that node would flow across the constrained line.

Changes in generation at other nodes will have less impact on the constrained line.

This implies that the marginal cost at a node could be thought of as depending on the marginal cost (per

MWh) of the constraint, multiplied by the proportion of a 1 MW increase in demand at that node that would

actually flow over the constraint. Transmission losses also lead the marginal cost of power to differ at every

point on the system. As electricity flows through the network, an amount proportional to the current

squared is lost in heating the wires, and there are also losses in transformers and other pieces of equipment.

Because the heating losses depend upon the current squared, the marginal loss is twice the average loss.

This implies that the marginal cost of meeting a demand at the importing end of a heavily loaded line can

be significantly greater than at the exporting end.

At the winter peak, it could be necessary to generate 106 MW in the north of England in order to meet an

extra 94 MW of demand in the south-west, if no closer plants were available.

If the system has been dispatched in a way that minimizes overall costs, then the marginal cost at any node

where a generator is running with spare capacity is the marginal cost of that generator. However, we would

obtain the same number if we calculated the marginal cost of generation at some arbitrary point on the

network (known as the ‘swing bus’), plus the cost in terms of losses of getting the power from the swing bus

to our location, plus the impact of additional demand at our location on constraints. The same formula is

used to calculate the marginal cost of meeting demand at all other nodes on the system.

10

EXAMPLE The following graph represents the optimal generation mix.

Annualized capital cost

[€/MW year]

Marginal cost

[€/MWh]

Peak-load power plant 30 000 35

Base-load power plant 100 000 25

VOLL (Value Of Lost Load) 95

Price Cap 45

Market equilibrium for different levels of demand can be represented as follows:

11

In long-run the price cap has major effects on the optimal mix, as can be seen from the graph:

Considering the price cap and the capacity payment, the situation changes again:

12

Adding a wind power plant, it becomes:

Results can be summed up in the following table:

SCENARIO

No price cap Price cap No price cap

Wind energy

Price cap

Wind energy

Base-load power plant 3 274 3 274 1 609 1 609

Peak-load power plant 3 633 2 233 4 418 2 383

Total capacity 6 907 5 507 6 027 3 992

Maximum demand 7 961 7 961 7 961 7 961

It is clearly evident that the price cap causes a loss of:

1400 MW of peak capacity in case of absence of wind energy

2035 MW of peak capacity in case of presence of wind energy

13

CAPACITY REMUNERATION MECHANISM (CRM)

MOTIVATION As long as demand is sufficiently price responsive, the interaction between available capacity and demand

determines the economically optimal level of installed capacity through the prices established in the market.

The level of adequacy is therefore determined by the market.

The political acceptability of the adequacy provided by energy-only markets depends on the frequency with

which prices reach very-high levels, and the political implications of such high prices.

It is the political unacceptability of extreme prices in energy-only markets that push MSs to intervene (e.g.

by introducing price caps in order to reduce the level of price spikes). Hence the need for Capacity

Remuneration Mechanisms (CRMs).

CLASSIFICATION A variety of CRMs have been proposed. They can be classified according to whether they are volume-based

or price-based. Volume-based CRMs can be further grouped in targeted and market-wide categories.

STRATEGIC RESERVE

In a Strategic Reserve scheme, some generation capacity is set aside to ensure security of supply in

exceptional circumstances, which can be signalled by prices in the day-ahead, intra-day or balancing

markets increasing above a certain threshold level.

An independent body, for example the Transmission System Operator, determines the amount of capacity

to be set aside to achieve the desired degree of adequacy and dispatches it whenever due.

The capacity to be set-aside is procured and the payments to this capacity determined through a (typically

year-ahead) tender and the costs are borne by the network users.

Ca

pa

city

re

mu

ne

rati

on

me

cha

nis

ms

Volume based

TargetedStrategic

reserve

Market-wide

Capacity obligation

Reliability option

Capacity auctionPrice based

Capacity payment

14

CAPACITY OBLIGATION

A Capacity Obligation scheme is a decentralized scheme where obligations are imposed on large

consumers and on load serving entities (LSE, also referred to as suppliers/retailers), to contract a certain

level of capacity linked to their self-assessed future (e.g. three years ahead) consumption or supply

obligations, respectively.

The capacity to be contracted is typically higher, by a reserve margin determined by an independent body,

than the level of expected future consumption or supply obligations.

The obligated parties can fulfil their obligation through ownership of plants, contracting with

generators/consumers and/or buying tradable capacity certificates (issued to capacity providers).

Contracted generators/consumers are required to make the contracted capacity available to the market in

periods of shortages, defined administratively or by market prices rising above a threshold level. Failure to

do so may result in penalties.

A (secondary) market for capacity certificates may be established, to promote the efficient exchange of

these certificates between generators/consumers providing capacity and the obligated parties or between

obligated parties.

CAPACITY AUCTION

Capacity Auction scheme is a centralised scheme in which the total required capacity is set (several years)

in advance of supply and procured through an auction by an independent body.

The price is set by the forward auction and paid to all participants who are successful in the auction. The

costs are charged to the suppliers who charge end consumers

Contracted capacity should be available according to the terms of the contract.

RELIABILITY OPTIONS

Reliability Options (ROs) are instruments similar to call options. According to them, contracted capacity

providers (typically generators) are required to pay the difference between the wholesale market price (e.g.

the spot price) and a pre-set reference price (i.e. the strike price), whenever this difference is positive, i.e.

the option is exercised. In exchange, they receive a fixed fee, thus benefitting from a more stable and

predictable income stream.

Under a RO scheme, the incentive for the contracted generator to be available (at times of scarcity) arises

from the high market price and from the fact that, if not available and therefore not dispatched, it will have

to meet the payments under the RO without receiving any revenue from the market.

The holders of ROs effectively cap their electricity purchase price at the level of the strike price, since

whenever the market price increases above this level, the excess will be “reimbursed” through the payment

made under the ROs.

A scheme based on ROs usually rests on an obligation imposed on large consumers and on suppliers to

acquire a certain amount of ROs, linked to their (self-assessed) future consumption or supply obligations,

respectively.

15

Different RO variants can be designed, depending on whether the scheme is purely financial or also involves

an obligation to have and make capacity available when the option is exercised (or otherwise face a penalty);

in this latter case the RO scheme becomes similar to a scheme based on Capacity Obligations.

CAPACITY PAYMENTS

Capacity Payments represent a fixed price paid to generators/consumers for available capacity. The

amount is determined by an independent body. The quantity supplied is then independently determined by

the actions of market participants.

ACTUAL SITUATION IN EUROPE