POST IMPLEMENTATION REVIEW OF P305 - ELEXON · BSC Modification P305 introduced a number of changes to the calculation of the cash-out price, and was implemented on 5 November 2015

POST IMPLEMENTATION REVIEW OF P305

EXEC SUMMARY AND CONTENTS

This document summarises data and analysis provided by ELEXON to contribute to BSC Parties’ understanding

of BSC Modification P305 ‘Electricity Balancing Significant Code Review Developments’.

Using BSC data and considering the six month period after P305 was implemented, the following can be

observed:

● Overall, the market was more long than short since the implementation of P305 – the system was

net long in 62% of Settlement Periods since the introduction of Modification P305 (compared to

57% of Settlement Periods in the same time period of the last year);

● Parties’ Imbalance Volumes in the six months following P305 were the greatest compared to the

same period in the last four years;

● System Prices have decreased on average but a greater number of incidents of more extreme

System Prices have been seen;

● Parties’ Trading Charges have increased following the implementation of P305, but by a small

amount (£2/MWh per day) for most Parties;

● The Reserve Scarcity Price (RSP), Demand Control actions or Contingency Balancing Reserve

actions have not been used since the implementation of P305 and therefore it was not possible to

fully assess the impact of these parameters.

CONTENTS

EXEC SUMMARY AND CONTENTS .............................................................................................................. 1

1. BACKGROUND AND SCOPE OF THE ANALYSIS ................................................................................ 2

2. BALANCING BEHAVIOUR ............................................................................................................... 11

3. OVERVIEW OF PRICE TRENDS ...................................................................................................... 19

4. IMPACT ON PARTIES TRADING CHANGES .................................................................................... 29

5. PARAMETER ANALYSIS ................................................................................................................. 34

LIST OF APPENDICES .............................................................................................................................. 42

GLOSSARY ............................................................................................................................................... 43

CLASSIFICATION OF PARTY TYPES ......................................................................................................... 45

SUMMARY OF DATA ISSUES FOLLOWING THE IMPLEMENTATION OF P305 ......................................... 48

PUBLIC

Post Implementation Review

Page 1 of 33 v2.0 © ELEXON 2016

https://www.elexon.co.uk/mod-proposal/p305/


Post Implementation

Review

Page 2 of 49 V2.0 © ELEXON 2016

1. BACKGROUND AND SCOPE OF THE ANALYSIS

The BSC Panel, made up of Industry representatives, highlights annually its strategic priorities and how to

address these. This year a priority was P305 and as such the Panel asked ELEXON to monitor the impacts of

P305.

This document summarises data and analysis provided by ELEXON to contribute to BSC Parties’ understanding

of BSC Modification P305 ‘Electricity Balancing Significant Code Review Developments’. Our review compiles

relevant evidence using Settlement data, but we do not provide an assessment of the Modification.

BSC Modification P305

BSC Modification P305 introduced a number of changes to the calculation of the cash-out price, and was

implemented on 5 November 2015. It was raised to progress the conclusions to Ofgem’s Electricity Balancing

Significant Code Review (EBSCR), which looked at addressing the Authority’s concerns with electricity

balancing arrangements.

The System Sell Price (SSP) and System Buy Price (SBP) are the ‘cash-out’ or ‘Energy Imbalance’ prices that

are used to settle the difference between contracted generation or consumption and the amount that was

actually generated or consumed in each half hour trading period. P305 implemented the following changes:

● A reduction in the Price Average Reference (PAR) value to 50MWh and the Replacement PAR

(RPAR) value to 1MWh upon implementation, and reduce the PAR value further to 1MWh on 1

November 2018;

● A single imbalance price so that SSP and SBP are equal to each other in each Settlement Period;

● A price for Short Term Operating Reserve (STOR) actions using a Reserve Scarcity Price (RSP)

which is determined with reference to a ‘static’ Loss of Load Probability (LoLP) function upon

implementation before switching to a ‘dynamic’ function on 1 November 2018; and

● A price for Demand Control actions at Value of Lost Load (currently £3,000/MWh) and a process

for correcting participants’ imbalance volumes following such an event.

National Grid raised P305 on 30 May 2014. The Panel agreed to submit P305 to an Assessment Procedure,

during which it was issued for industry Impact Assessment and the Workgroup’s Assessment Procedure

Consultation. The Workgroup recommended that the Alternative Modification should be approved, and its

Assessment Report was presented to the Panel on 12 February 2015.

The Panel initially recommended that both the Proposed and Alternative Modifications should be rejected, and

issued P305 for its Report Phase Consultation. The Panel made its final recommendation that both the

Proposed and Alternative Modifications should be rejected at its meeting on 12 March 2015.

The Authority approved the P305 Proposed Modification on 2 April 2015 for implementation on 5 November

2015 as part of the November 2015 Release.

The decision to approve BSC Modification P305 was based on defects identified in the calculation of cash-out

prices by the Electricity Balancing Significant Code Review (EBSCR). The concerns were as follows:

● Cash-out prices were calculated using an average of the cost of the actions that the System

Operator (SO) takes to balance the system, rather than the marginal action;

● The ‘dual price’ arrangements created unnecessary imbalance costs for parties because the price

for ‘helpful’ imbalances did not reflect the savings these imbalances create for the SO;

● They excluded the costs borne by the consumer during disconnection and voltage reduction; and



Post Implementation

Review


● The method for pricing reserve costs into cash-out did not accurately reflect the real time value of

this reserve, and excluded the cost of some reserve products altogether.

The EBSCR Final Policy Decision concluded that these defects could increase the cost of ensuring security of

supply to consumers because it could lead to inefficient balancing and dampen incentives for the market to

provide flexibility.

Analysis provided

We have carried out analysis using BSC data1 focusing on the six month period following the implementation of

P305 (i.e. 5 November 2015 to 30 April 2016).

The intention is to look back at the period since implementation and analyse what System Prices and Trading

Charges would have been had P305 not been introduced, and to look forward ahead of ‘phase two’ of the

changes being introduced in November 2018. We also make some comparison of the same six month period

across different years. There is one important caveat to any analysis, and that is that we cannot account for

any behavioural change that may have resulted from different System Prices.

To facilitate the analysis, we compare three pricing scenarios:

1. Live Prices (“Live”) – this is the price calculation that is used in calculating System Prices since 5

November 2015.

2. P217A Price Scenario (“P217”) – this is the price calculation that applied before P305 was

implemented on 5 November.

3. November 18 or Price Scenario (“Nov-18”) – this is the ‘phase two’ price calculation that will

apply from 1 November 2018, specifically to calculate imbalance prices with a lower Price Average

Reference (PAR) value and a Value of Lost Load (VoLL) of £6,000/MWh.

We also assess the impact of Parties’ Trading Charges during the period. To carry out this analysis, we have

split different BSC Parties into different Party Types2:

● Independent Generator;

● Non Physical Trader;

● Renewable Generator;

● Vertically Integrated Player; and

● Independent Supplier.

We welcome your comments on the contents of this report. Please email any feedback to [email protected].

1 The data used is a combination of Settlement Runs using at least ‘SF’ data. 2 Note that the Party Grouping is based on our latest information and best understanding of Parties. See

Appendix 2 for details.

mailto:[email protected]


Post Implementation

Review


Summary of workgroup views and evidence provided

We have aimed to provide data to address the views of the Modification workgroups. A summary of views

provided against the BSC Objectives are provided below. For a full report of the workgroup and the Panel’s

views on P305, see the Final Modification Report3.

Summary of Workgroup Members’ Views4

BSC Objectives Workgroup members’ views Evidence Provided Out of scope

(a) The efficient

discharge by the

Transmission

Company of the

obligations imposed

upon it by the

Transmission

Licence

Neutral (unanimous) N/A N/A

(b) The efficient,

economic and co-

ordinated operation

of the National

Electricity

Transmission

System

Beneficial (minority)

Strengthens incentive to balance

efficiently, particularly in times of

tight margin

Potential increase in liquidity

which will help participants

balance ahead of Gate Closure

Detrimental (majority)

LoLP values could send out false

signals and could encourage

balancing after Gate Closure if

high

Volatile prices may cause

participants to take longer

positions to avoid the

consequences of being short

More marginal prices increases

the risk of balancing actions

incorrectly impacting the

Absolute

Imbalance

Volumes

Party Imbalance

Volumes

Market Index

Volumes

Aggregated Party

Imbalances

Improvements in cost-

reflectivity will encourage

investment, driving long

run cost savings

Better reflects the value of

flexible generation, which

may help defer the

decommissioning of such

plant

3 The Final Modification Report (FMR) can be found at: https://www.elexon.co.uk/mod-proposal/p305/ 4 This shows a summary the different views expressed by Workgroup members – not all members necessarily agreed with all of these

views.



Post Implementation

Review


imbalance price in subsequent

Settlement Periods

(c) Promoting effective

competition in the

generation and

supply of electricity

and (so far as

consistent

therewith)

promoting such

competition in the

sale and purchase

of electricity

Beneficial (minority)

Allows flexible and reliable plant

to gain advantage that reflect

their value to consumers

Single price removes the

inefficient price spread and the

net imbalance costs that creates

Incentivises participants to

balance positions, increasing

liquidity and encouraging

investment in flexible capacity

Sharpens the signals of scarcity

to the market

Detrimental (majority)

Volatile prices will have a

detrimental effect on smaller

participants

The distributional effects of P305

are unknown

The reduction in PAR to 50MWh

is too large a step and the

impacts this will have are

unknown

Single price may result in less

trading, reducing liquidity

Party Imbalance

Volumes

Market Index

Volumes

System Prices

Improves incentives for

flexible and reliable plant

to enter the market

(d) Promoting

efficiency in the

implementation of

the balancing and

settlement

arrangements

Detrimental (minority)

Introduces complex processes

with little proven benefit

Neutral (majority)

N/A N/A

(e) Compliance with

the Electricity

Regulation and any

relevant legally



Post Implementation

Review


binding decision of

the European

Commission and/or

the Agency [for the

Co-operation of

Energy Regulators]

(f) Implementing and

administrating the

arrangements for

the operation of

contracts for

difference and

arrangements that

facilitate the

operation of a

capacity market

pursuant to EMR

legislation



Post Implementation

Review


Market context

Outturn System Prices in the six months since the implementation of P305 are as much a function of the

market fundamentals as the changes to the calculation that were introduced; and this is an important caveat

to any historical comparisons.

It may be that the market conditions since the implementation of P305 have meant that some of the reforms

have not had the impact that some stakeholders expected. For example, in National Grid’s Winter Outlook

Review5, they noted that:

● Winter 15/16 was one of the mildest winters in almost sixty years;

● Peak demand was 1GW lower than expected; and

● Operational conditions were less challenging than anticipated.

Our own analysis on the changes in market fundamentals is presented in the section below. Graphs 1 and 2

show the fuel mix for the first six months of P305, and how this compares to the same time period from the

preceding year, using metered output data from generating BMUs (embedded generation is therefore not

captured).

Metered output from gas-fired plant increased by

over 10%, while coal output fell around 15%. Fuel

prices were largely the driving force behind this

trend, with gas plants becoming more economical to

run than coal plants. Wind also saw an increase;

with increased capacity installed and higher levels of

wind over 2015/16 compared to the previous year.

The average available generation per Settlement

Period has fallen since Winter 2014/15. This can be

seen in Graph 3, which shows average installed

capacity margins over time using Effective

Maximum Export Limits. Some of the factors

causing this are plant closures, decreasing demand

and increasing embedded generation. Available

capacity shows a seasonal shape, with less capacity

5 http://media.nationalgrid.com/media/1293/ng-winter-review-2016.pdf

Graph 3 – Average installed capacity margins by over time using Effective Maximum Export Limits

20

22

24

26

28

30

32

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

2014 2015 2016

GW

h

Graph 2 - 2014/15 Fuel Mix

37%

17%

22%

9%

8%5% 2% Gas

Coal

Nuclear

Wind

Interconnector

Bio

Hydro

Oil

Graph 1 - 2015/16 Fuel Mix

26%

32%

21%

7%

8%5% 1%

Gas

Coal

Nuclear

Wind

Interconnector

Bio

Hydro

Oil

http://media.nationalgrid.com/media/1293/ng-winter-review-2016.pdf


Post Implementation

Review


available over the summer months, as

generators typically performing maintenance in

summer.

Graph 4 shows average accepted Bid Price by

fuel type. Wind generators typically submit

negatively-priced bids to offset the cost of the

Renewable Obligation Certificates that they will

not receive if they are not generating. The

average Bid price of coal has tended

downwards, from around £40/MWh in January

2014 to around £20/MWh in April 2016.

Graph 5 shows average Offer Price by fuel type

(wind has been excluded since this fuel type

rarely has substantial Offers accepted).

The period since implementation of P305 was

also a period of continued falling wholesale

prices. Graph 6 shows market prices, as

measured by the Market Index Price (MIP). The

MIP is a price (expressed in £/MWh), calculated

for each Settlement Period, to reflect the price

of wholesale electricity in the short term or

intra-day markets. Prices and volumes for

trades that occur within 12 hours of Gate

Closure are submitted by the Market Index Data

Provider(s) (currently APX and N2EX) and these

are used to calculate the Market Index Price6.

Aside from the impact that these conditions

may have had on the pricing of the underlying

balancing actions (such as Bids and Offers in the

Balancing Mechanism) used to set the price they

have also had a more direct impact on cash-out

prices. There were no Contingency Balancing

Reserve or Demand Control actions taken since

the November release, both of which would be

priced at £3,000/MWh for the purposes of the

System Price calculation. Further, the ‘less tight’

margins meant that the Reserve Scarcity Price

(RSP) did not have any impact on System Prices.

6 For full detail of the Market Index Price, see the Market Index Definition statement here:

https://www.elexon.co.uk/wp-content/uploads/2012/01/mids_v7.0.pdf

Graph 6 – Average Market Index price by day

0

5

10

15

20

25

30

35

05/1

1/2

014

28/1

1/2

014

21/1

2/2

014

13/0

1/2

015

05/0

2/2

015

28/0

2/2

015

23/0

3/2

015

15/0

4/2

015

08/0

5/2

015

31/0

5/2

015

23/0

6/2

015

16/0

7/2

015

08/0

8/2

015

31/0

8/2

015

23/0

9/2

015

16/1

0/2

015

08/1

1/2

015

01/1

2/2

015

24/1

2/2

015

16/0

1/2

016

08/0

2/2

016

02/0

3/2

016

25/0

3/2

016

17/0

4/2

016

MIP

(£

/M

Wh

)

Total Linear (Total)

Graph 4 – Average accepted Bids over time, split by fuel type

-120

-100

-80

-60

-40

-20

0

20

40

60

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

2014 2015 2016

Bid

Pri

ce

(£

/M

Wh

)

Coal Hydro Pumpedstorage

Wind Gas/CCGT/OCGT

Graph 5 - Average accepted Offers over time, split by fuel type

0

20

40

60

80

100

120

140

160Ja

n

Mar

May

Jul

Sep

Nov

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

2014 2015 2016

Off

er

Pri

ce

(£

/M

Wh

)

Coal Hydro Pumpedstorage Gas/CCGT/OCGT

https://www.elexon.co.uk/wp-content/uploads/2012/01/mids_v7.0.pdf


Post Implementation

Review


Related work

This review focuses on the impact of the changes introduced as part of P305; other reviews are performed in

relation to specific aspects of the arrangements.

Market Index Price Review and issue group

The Market Index Definition statement is the document which contains the methodology for determining the

Market Index Price (MIP). Before P305, it was used to set one of the Imbalance Prices in each half hour.

Following the implementation of P305, it is only used in rare, defaulting circumstances (e.g. when there are no

energy balancing actions taken in a half-hour).

At its October 2015 meeting, the BSC Panel, following the Imbalance Settlement Group’s (ISG)

recommendations, asked ELEXON to form an Issue Group to look at the use of the MIP in those defaulting

situations. The Issue Group will be formed in June 2016 and will consider whether the MIDs remains fit for

purpose, and whether there are other options for setting default prices. The Issue Group’s recommendations

will feed into the annual MIDS review in August 2016, and the outcome will be discussed with the ISG and the

Panel.

Continuous Acceptance Duration Limit (CADL) and De-Minimis Threshold

The Continuous Acceptance Duration Limit (CADL) and the De-Minimis Acceptance Threshold (DMAT) are

parameters that are used in the calculation of the System Price, and they are subject to review every two

years. The next review of these parameters is due in October 2016.

CADL is used to identify short-duration Bid Offer Acceptances (BOAs) that are most likely to be associated with

system balancing actions and potentially exclude them from the Main Energy Imbalance Price calculation. It

has been set at 15 minutes since its introduction in 2001.

DMAT removes balancing actions smaller than a set value, currently 1MWh, from the Main Energy Imbalance

Price calculation.

Value of Lost Load (VoLL) Process Review

Value of Lost Load (VoLL) is a defined parameter in the BSC, and is currently set to £3,000/MWh. P305 has set

the VoLL at this level until 1 November 2018, when the VoLL will be increased to £6,000/MWh. As the VoLL

has been planned to be increased on this date, no review has currently been scheduled.

ELEXON plans to carry out a review of the VoLL in 2019. Subject to BSC Panel approval, the review process will

include bringing a consultation paper to the ISG for approval, reporting industry views back to the ISG and

presenting the ISG’s recommendations to the Panel. Any change would need a BSC Modification and would

need the Authority’s approval. In any case the Authority would receive a report from the Panel with its

decision.

Loss of Load Probability (LoLP) Calculation Statement review

The Loss of Load Probability Calculation Statement7 is a document which sets out the methodology for

determining the relationship between De-Rated Margin and the probability that there will be loss of load for a

given Settlement Period. LoLP is used to determine the Reserve Scarcity Price for a Settlement Period. Section

1.5 of the LoLP Statement specifies that the Panel may review this Statement from time to time and make

changes, subject to the Authority’s approval, in accordance with BSC Section T 1.6A. This section requires that

any suggested changes to be submitted to Industry for consultation and that the Transmission Company, as

well as all BSC Parties, to be informed of any change approved by the Authority.

7 https://www.elexon.co.uk/wp-

content/uploads/2014/10/37_244_11A_LOLP_Calculation_Statement_PUBLIC.pdf

https://www.elexon.co.uk/bsc-related-documents/balancing-settlement-code/bsc-sections/

https://www.elexon.co.uk/wp-content/uploads/2014/10/37_244_11A_LOLP_Calculation_Statement_PUBLIC.pdf



Post Implementation

Review


As stated in Section T 1.6A.7, the Panel may delegate responsibility to maintain and review the LoLP

Statement. Subject to BSC Panel’s approval, ELEXON’s review process will be overseen by the ISG and involve

Industry consultation before presentation to the Panel for decision. The Authority would need to approve any

change to the LoLP Statement. The LoLP calculation will change to a Dynamic LoLP Function on 1 November

2018, therefore, there is currently no plan to perform a review of the LoLP Statement. Industry will be

informed if the Panel determines that a review is necessary.

Review of Credit Modifications

Since November 2014, a number of BSC Modifications have been introduced to improve the Credit Cover

process: P306 ‘Expanding the definition of a ‘Letter of Credit’ to include regulated insurance companies’, P307

‘Amendments to Credit Default arrangements’ and P310 ‘Revised Credit Cover for Exporting Supplier BM Units’.

A first stage Post Implementation Review of those Modifications was undertaken in January (ISG177/02) and

February (Panel 249/15) 2016 and a second stage will be presented at the July 2016 ISG meeting and August

2016 Panel meeting.





https://www.elexon.co.uk/meeting/isg-177/

https://www.elexon.co.uk/meeting/bsc-panel-249/


Post Implementation

Review


2. BALANCING BEHAVIOUR

In this section, we present data on:

● The overall imbalance on the system since P305 was introduced, and how this compares to

imbalances in the same time period the previous year (“Net Imbalance Volume”)

● Parties’ overall imbalances across the period, how these have changed over time, and how they

compare to the same time period in the previous year (“Parties aggregated Imbalances”)

● Parties’ overall imbalances, and how these differ at different times (peak times against off-peak)

● A measure of volumes of trades in the intra-day forward markets, compared to the same period

last year (“Traded volumes from Market Index Data Providers”)

Market balancing as measured by Net Imbalance Volume (NIV)

There were mixed views about the impact that

P305 would have on the market’s incentive to

balance. Some argued that it would increase

incentives to balance efficiently (noting that this

would not necessarily mean always having a fully

balanced position ahead of Gate Closure), while

others argued that it would increase incentives to

go long, in response to price volatility or to avoid

the consequences of being short.

Net Imbalance Volume (NIV) is the net of all

balancing actions taken by the System Operator for

a Settlement Period. It indicates whether the

system was overall long or short in a Settlement Period, and this can be used as one measure of overall

imbalance or length of the system. Positive NIV

denotes a short system and vice versa.

Graph 7 shows the proportion of Settlement

Periods that were long and short and how this

compares to the same time period from the

preceding year. Overall, the market has been long

more frequently in 2015/16 – the net imbalance on

the system was long in 62% of Settlement Periods

from November 2015 to April 2016, compared to

57% in the same period for the preceding year.

However, considering this against the previous four

years the trend seems less pronounced – Graph 8

shows the proportion that the system was net long

by month in November 2015 to April 2016, and

how this compares to the same period across the

previous years.

20%

30%

40%

50%

60%

70%

80%

90%

Nov Dec Jan Feb Mar Apr

% o

f lo

ng

pe

rio

ds

12/13 13/14 14/15 15/16

Graph 8 - Settlement Periods with a long system

Graph 7 – Average System length per Settlement Period by Month

0%10%20%30%40%50%60%70%80%90%

100%

14/1

5

15/1

6

14/1

5

15/1

6

14/1

5

15/1

6

14/1

5

15/1

6

14/1

5

15/1

6

14/1

5

15/1

6


Pe

rce

nta

ge

of

Se

ttle

me

nt

Pe

rio

ds

Long system Short system


Post Implementation

Review


Graph 9 shows how long and short the market has been on average across the Settlement Day over 2014/15

(5 November 2014 – 30 April 2015) and Graph 10 shows the same data for the 2015/16 year (5 November

2015 – 30 April 2016). The system length appears to vary more across the day in 2014/15, with the proportion

of times the system was long greater over the morning peak (Settlement Periods 13 to 20) and the evening

peak (Settlement Periods 35 to 43).

Graph 10 – Average System length per Settlement Period 2015/16

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47

Pe

rce

nta

ge

of

Se

tte

lme

nt

Pe

rio

ds

Sum of Long Sum of Short

Graph 9 – System length per Settlement Period 2014/15

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47

Pe

rce

nta

ge

of

Se

ttle

me

nt

Pe

rio

ds

Sum of Long Sum of Short


Post Implementation

Review


Graph 11 shows the magnitude of NIVs observed in 2015/16 compared to 2014/15 (positive NIVs denote a

short system and negative NIVs denote a long system).

Graph 11 – Frequency of Net Imbalance Volumes (NIVs)

0

200

400

600

800

1000

1200

-20

00

- -

19

00

-18

00

- -

17

00

-16

00

- -

15

00

-14

00

- -

13

00

-12

00

- -

11

00

-10

00

- -

90

0

-80

0 -

-7

00

-60

0 -

-5

00

-40

0 -

-3

00

-20

0 -

-1

00

0 -

10

0

20

0 -

30

0

40

0 -

50

0

60

0 -

70

0

80

0 -

90

0

10

00

- 1

10

0

12

00

- 1

30

0

14

00

- 1

50

0

16

00

- 1

70

0

18

00

- 1

90

0

20

00

-

Fre

qu

ency

NIV (MWh14/15 15/16


Post Implementation

Review


Market balancing as measured by Parties’ aggregated Imbalance

Absolute Imbalances

NIV shows the net length of the system

(i.e. it shows whether the system is long

or short for a given half-hour), but this

hides the long and short imbalances faced

by market participants in each half hour.

Party imbalances are the difference

between contracted volumes of energy

and physical production and consumption.

In Graph 12 we present data on Parties’

absolute Imbalance Volumes. A second

plot shows Energy Imbalance Volumes as

a percentage of total demand. This shows

that there has been an increase in both

absolute Imbalance Volumes and

Imbalance Volumes as a proportion of

total demand over the past two years.

This trend appears to have continued

since the implementation of P305.

Graph 13 presents absolute Imbalance

Volume data from November to April for

the past 4 years, aggregated by month.

Parties’ imbalance volumes were higher in

15/16 for every month, except in April

14/15 when Imbalance Volumes were

0.07 TWh higher.

Graph 14 shows average absolute

Imbalance Volumes by Settlement Period

in 15/16 compared to the same period

from the previous year. The absolute

imbalance volumes are higher in all

Settlement Periods in 2015/16.

Graph 13 – Absolute imbalance volumes by year

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2


Imb

ala

nce

Vo

lum

e (

TW

h)

12/13 13/14 14/15 15/16

Graph 12 – Absolute Imbalance Volumes and as a percentage of total demand by month

-

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

0

0.5

1

1.5

2

2.5

Apr-

14

Jun-1

4

Aug-1

4

Oct

-14

Dec-

14

Feb-1

5

Apr-

15

Jun-1

5

Aug-1

5

Oct

-15

Dec-

15

Feb-1

6

Apr-

16

Imb

ala

nce

Vo

lum

e a

s a

Pe

rce

nt

of

To

tal D

em

an

d

Ab

so

lute

Im

ba

lan

ce

Vo

lum

e

(TW

h)

Imb Vol Imb Vol as percent of Total Demand

Graph 14 – Absolute imbalance volumes by Settlement Period

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

1 3 5 7 9 11131517192123252729313335373941434547

Imb

ala

nce

Vo

lum

e (

MW

h)

14/15 15/16


Post Implementation

Review


Party Imbalances by Long and Short

Next we present Party imbalances split by

whether they were long or short imbalances.

Some stakeholders expressed the view that

P305 may create incentives to ‘go long’ (i.e.

purposely contract more, demand less or

produce more electricity) particularly over the

peak periods, or in response to price

volatility. Overall, Parties long imbalance over

5 November 2015 to 30 April 2016 were 25%

greater and short imbalances 19% greater

compared to the same period from the

previous year.

Graph 15 shows Party imbalances by long

and short imbalances per month in 2015/16,

and how these compared to the same period from the previous year (long balance volumes are displayed as

positive volumes and short imbalance volumes are displayed as negative volumes). Long imbalances were

greater in each month in 2015/16, with the exception of April 2016. Short imbalances were greater in every

month in 2015/16.

Table 1 shows long and short imbalance volumes split by peak8 and off-peak Settlement Periods for 5

November 2015 until 30 April 2016 and the same period for 2014/15. The average imbalance volume was

greater (both longer and shorter)

over the peak Settlement Periods

– in 2015/16 long imbalance

volumes over the peak were 31%

greater than off-peak long

imbalance volumes, and short

imbalance volumes over the peak

were 27% greater than off-peak

short imbalance volumes.

8 Peak periods are defined as Settlement Periods 15 to 38 for the purposes of this analysis.

Table 1 – Aggregated long and short party imbalances by peak and off-peak. Averaged by Settlement Period

Average Long

Volume (MWh)

Average Short

Volume (MWh)

Average Long

Volume (MWh)

Average Short

Volume (MWh)

Off-peak 5,908 5,196 7,415 6,071

Peak 8,000 6,496 9,734 7,690

14/15 15/16

Graph 15 - Aggregated long and short party imbalances in 2015/16 with percentage change from previous year

+9%

+48%

+22%+43%

+45%-10%

+29% +17% +18% +16% +5% +30%-20%

-10%

0%

10%

20%

30%

40%

50%

60%

-15

-10

-5

0

5

10

15

20

Nov Dec Jan Feb Mar AprA

cco

un

t Im

ba

lan

ce

(T

Wh

)


Post Implementation

Review


Table 2 presents imbalance volumes as a percent of Total Demand by different Party Type as categorised

using BSC Party IDs (see appendix two for classification of Party types for further detail). It can be seen that

independent Generators, Non-Physical Traders and Renewable Generators long imbalance percentages

increased significantly when comparing 2014/15 to 2015/16. This trend is being driven by a small number of

Parties within each Party type. The short imbalances remain largely consistent over the two time periods, the

only exception being non-physical traders, whose short volumes more than doubled. Again this trend is being

driven by a small number of market participants.

Date Independent Generator Non Physical Trader Renewable Generator Supplier Vertically Integrated

14/15 0.072% 0.052% 0.043% 0.798% 1.406%

15/16 0.177% 0.188% 0.073% 1.164% 1.350%

Date Independent Generator Non Physical Trader Renewable Generator Supplier Vertically Integrated

14/15 0.125% 0.032% 0.036% 0.427% 1.337%

15/16 0.113% 0.077% 0.042% 0.498% 1.562%

Long

Short

Table 2 – Aggregated Imbalance Volume (MWh) as a percent of Total Demand by Party Type


Post Implementation

Review


Balancing production and consumption accounts

Parties have two energy accounts – a

production account and a consumption

account – and Imbalance Volumes are

calculated separately for these.

Under a dual cash-out price, long and short

imbalances were subject to different System

Prices, and the spread between the two

prices created a greater incentive for Parties

to balance each of their energy accounts

separately. However, since the introduction

of a single cash-out price as part of P305,

any opposing imbalances will have the same

price, so the difference nets off, and the incentive has lessened.

Graph 16 shows how daily gross Imbalances Volumes have changed over time. Following October 2015, both

the production and consumption accounts have been on an upward trajectory, suggesting greater Imbalance

Volumes.

Graph 16 – Daily gross imbalance volumes by account

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

2014 2015 2016M

Wh

Production Consumption


Post Implementation

Review


P305 and intra-day liquidity The P305 Workgroup expressed mixed

feelings as to whether P305 would have a

beneficial or detrimental impact on liquidity

in the wholesale market.

Graph 17 shows Market Index Volumes

(MIV) as a measure of liquidity in the spot

markets from November 2015 to April 2016,

compared to the same time period of the

preceding year. These are volumes of trades

submitted by the Market Index Data

Provider(s) and reflect how much trade

occurred for the weighted products as

defined in the Market Index Definition

Statement (MIDS). They do not, therefore,

show total market liquidity, but can be used to give an indication of the level of trading before Gate Closure.

Using this metric, traded volumes have not changed significantly, with average volumes per day over the

2014/15 period as 35,587 MWh and 35,151 MWh over the 2015/16 period.

Graph 17 –Market index data traded volumes by day

0

10000

20000

30000

40000

50000

60000

70000

05/1

1/2

014

19/1

1/2

014

03/1

2/2

014

17/1

2/2

014

31/1

2/2

014

14/0

1/2

015

28/0

1/2

015

11/0

2/2

015

25/0

2/2

015

11/0

3/2

015

25/0

3/2

015

08/0

4/2

015

22/0

4/2

015

Vo

lum

e (

MW

h)

2014/15 2015/16


Post Implementation

Review


3. OVERVIEW OF PRICE TRENDS

This section provides an overview of imbalance prices in the 6 months since the implementation of P305, as

well as prices recalculated using two different pricing scenarios:

● The P217 Price Scenario – the price calculation that was used before P305 was implemented.

● The November 2018 Price Scenario – the price calculation that will be used from 1 November

2018

For historic analysis we compare prices when the System is Long (“Long System Price”) and Short (“Short

System Price”) to capture the impact of the changes that made the cash-out price ‘more marginal.’

Overview of System Prices since the implementation of P305

Graph 18 shows the monthly average System Prices since P305 was implemented, as well as System Prices

when the system was short and when the

system was long. Average long System

Prices have decreased since November 2015

and short System Prices have increased. The

average price has been between

approximately £35/MWh and £40/MWh.

Tables 3 and 4 shows an overview of

System Prices since 5 November 2015, split

by a long and short system.

0

10

20

30

40

50

60

70

80


2015 2016

Syst

em

Pri

ce

(£

/M

Wh

)

Long System Short System Average price

Graph 18 - Monthly average System Prices over time

Table 4 – Overview of long prices

Date Min (£/MWh) Max (£/MWh) Mean (£/MWh) Standard Deviation (£/MWh)

Nov-15 -39.96 40.00 27.27 6.65

Dec-15 -73.48 42.00 26.67 10.94

Jan-16 -35.00 43.29 27.12 6.78

Feb-16 -59.95 38.00 23.98 6.04

Mar-16 -63.02 39.70 21.92 10.03

Apr-16 -39.19 40.00 22.33 5.54

Table 3 – Overview of short prices

Date Min (£/MWh) Max (£/MWh) Mean (£/MWh) Standard Deviation (£/MWh)

Nov-15 18.43 178.00 53.15 19.69

Dec-15 39.00 140.00 53.53 15.33

Jan-16 33.17 225.00 57.24 27.40

Feb-16 30.00 151.00 54.92 21.67

Mar-16 21.10 517.55 71.66 59.27

Apr-16 30.46 161.19 59.20 30.41


Post Implementation

Review


System Prices since P305 implementation compared to 2014/15

This section looks at System Prices since

the implementation of P305 and

compares these to historic System Prices.

Any differences will be a result of a

combination of the changes to the cash-

out price calculation as well as any

changes to market fundamentals.

Graph 19 shows the average System

Price (averaging SSP and SBP in both

directions pre-305) from November 2014

to April 2016. Using this measure, the

average price of imbalance has fallen

since the introduction of P305 from

£41.70/MWh in October 2015 to

£36.10/MWh in April 2016.

Graph 20 shows the frequency of System Prices occurring within defined price bands in 2014/15 compared to

2015/16. In 2015/16, most System Prices were between £20 and £30/MWh, whereas in 2014/15 prices were

more frequently between £30 and £40/MWh.

Graph 20 – Frequency of System Prices

0

1000

2000

3000

4000

5000

6000

-10+

-10 -

0

0 -

10

10 -

20

20 -

30

30 -

40

40 -

50

50 -

60

60 -

70

70 -

80

80 -

90

90 -

100

100+

Fre

qu

en

cy

System Prices (£/MWh)

14/15 15/16

Graph 19 – Average System Price (long and short)

0

10

20

30

40

50

60

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

Feb

Mar

Apr

2014 2015 2016

Syste

m P

rice

(£

/M

Wh

)


Post Implementation

Review


Graphs 21 and 22 focus on the more extreme prices – above £70/MWh and less than £0/MWh. Eight

instances of prices greater than £250/MWh were seen in the 2015/16, whereas the highest price in 2014/15

was £173.71/MWh. The same trend is seen for the negative prices, with a lowest price of -£73.48/MWh in

2015/16 compared to -£57.26/MWh in 2014/15.

Graph 21 – Frequency of high System Prices

Graph 22 – Frequency of Low System Prices

0

20

40

60

80

100

120

70 -

80

80 -

90

90 -

100

100 -

110

110 -

120

120 -

130

130 -

140

140 -

150

150 -

160

160 -

170

170 -

180

180 -

190

190 -

200

200 -

210

210 -

220

220 -

230

230 -

240

240 -

250

250+

Fre

qu

en

cy


14/15 15/16

0

5

10

15

20

25

30

35

40

-80+

-80 -

-70

-70 -

-60

-60 -

-50

-50 -

-40

-40 -

-30

-30 -

-20

-20 -

-10

-10 -

-0.0

1

Fre

qu

en

cy


14/15 15/16


Post Implementation

Review


Comparison of System Prices using different price scenarios

This section compares live prices with two different pricing scenarios. First, we consider what prices would look

like with the P217 (pre-P305) price calculation to highlight the impact of P305. Before the implementation

of P305, the price calculation had:

● A PAR of 500MWh and an RPAR of 100MWh;

● No non-BM STOR volumes or prices included in the price stack;

● No RSP, and instead a Buy Price Adjuster (BPA) that recovers STOR availability costs; and

● No Demand Control, Demand Side Balancing Reserve (DSBR), or Supplementary Balancing Reserve

(SBR) actions priced at VoLL.

Graph 23 shows the average System Price when the system was short using the two calculations by

Settlement Period from November 2015 to April 2016. Graph 24 shows the same when the system was long.

Graph 23A and Graph 24A show the same price data displayed as a function of distribution.

0

5

10

15

20

25

30

35

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46

Syste

m P

rice

(£

/M

Wh

)

Settlement Period

Average of Live Price Average of P217

Graph 24 – Average long System Price

0

20

40

60

80

100

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46Syste

m P

rice

(£

/M

Wh

)

Settlement Period

Average of Live Price Average of P217

Graph 23 – Average short System Price

0

1000

2000

3000

4000

5000

-80 -

-70

-70 -

-60

-60 -

-50

-50 -

-40

-40 -

-30

-30 -

-20

-20 -

-10

-10 -

0

0 -

10

10 -

20

20 -

30

30 -

40

40 -

50

50 -

60

60 -

70

70 -

80

80 -

Fre

qu

en

cy

System Price (£/MWh)

Live P217

Graph 24A – Average short System Price

0

500

1000

1500

2000

10 -

20

30 -

40

50 -

60

70 -

80

90 -

100

110 -

120

130 -

140

150 -

160

170 -

180

190 -

200

Fre

qu

en

cy


Live P217

Graph 25A – Average long System Price


Post Implementation

Review


Graph 25 shows the magnitude of these price changes (there was no price change in 17.7% of Settlement

Periods).

On average, P305 prices were lower than P217 prices when the system was longer and higher when the

system was short. However there were some exceptions to this – 26.78% of live prices when the system was

short were lower than pre-305 prices. These differences were driven by the removal of the portion of the Buy

Price Price Adjuster (BPA) that was used to recover STOR availability fees. The proportion of the BPA related to

STOR availability fees was removed when the RSP was introduced with P305.

Graph 25 - P217 scenario System Price changes

0

500

1000

1500

2000

2500

3000

3500

4000

4500

-20+

-20 -

-15

-15 -

-10

-10 -

-5

-5 -

0

0 -

5

5 -

10

10 -

15

15 -

20

20 -

25

25 -

30

30 -

35

35+

Fre

qu

en

cy o

f P

rice

Ch

an

ge

System Price changes (£/MWh)

Long Short


Post Implementation

Review


Case study: the biggest price difference between the live and P217 scenarios when the system was short In this case study we compare the price stack under the current pricing arrangements with those under the

previous ‘P217A’ arrangements.

When the market is short, the System Price is set by Buy Actions (including Offers from the Balancing

Mechanism). However, we do not include all actions in the price calculation; there are a number of tagging

processes to remove some of actions from the volume of actions that are used to set the price (the “Priced

Volume”):

NIV tagging nets off the Buy and Sell actions to determine the overall direction of the system – the

Net Imbalance Volume (NIV). The smaller stack of actions is netted off the larger stack of balancing

actions, starting with the most expensive actions. Actions removed in this way are referred to as ‘NIV

Tagged’.

The energy imbalance price is calculated based on the volume-weighted average of a defined volume

of the most expensive actions remaining. This is the Price Average Reference Volume (PAR) and

before P305 it was a volume of 500MWh, but P305 reduced this volume to 50MWh.

The largest price difference between the live and P217a prices (when the market was short) occurred on 10

March 2016 at Settlement Period 40 and was driven by the reduction in the PAR volume. The live price was

£517.55/MWh, but under the old calculation the price would have been £272.31/MWh.

After NIV-tagging, there were 448MWh of Buy Actions left in the stack. Under the old price calculation, a full

average of these actions was taken (because the volume was less than the PAR of 500MWh). This volume was

made up of Buy actions from 20 different BMUs, with prices ranging from £40/MWh to £500/MWh. A weighted

average of these resulted in a price of £217.31/MWh (including a Buy Price Adjuster of £20.00/MWh)9. In the

live price scenario, however, the reduced PAR of 50MWh meant that only the most expensive 50MWh of

actions were averaged to set the price. In this case, this volume came from an Offer from a single BMU with a

price of £500/MWh (with the addition of a Buy Price Adjuster of £17.55).

9 This is added onto the price to account for National Grid’s option fees. See page 42 for further detail.

0

100

200

300

400

500

600

700

Priced Volume

(448MWh)

NIV tagged volume(Buys)

NIV tagged volume(Sells)

FULL STACK OF ACTIONS – 217A SCENARIO

-200

-100

Acti

on

vo

lum

e (

MW

h)

0

100

200

300

400

217A Price, PAR500 =

£272.31/MWh

Live Price, PAR50 =

£517.55/MWh*

PRICED VOLUME, PAR500 AND PAR50

*with the addition of a £17.55 Buy Price adjuster


Post Implementation

Review


November 2018 scenario

The November 2018 Scenario is intended to reflect the effect of changes to the imbalance price parameters

that are due to come in on 1 November 2018. These are:

● A reduction in the PAR value to 1MWh (RPAR will remain at 1MWh); and

● An increase in the VoLL to £6,000/MWh, which will apply to all instances of VoLL in arrangements,

including the RSP function.

A ‘dynamic’ LOLP function will also be introduced from November 2018, but we have been unable to capture

the impact of this in the pricing scenario.

Graphs 28 and 29 show System Prices from November 2015 to April 2016 recalculated using the November

2018 scenario. Prices are always lower when the system is long under the November 2018 scenario and higher

when the system is shorter. Graph 28A and Graph 29A show the differences between the live scenario and

the November 2018 scenario at Settlement Period level.

0

1

2

3

4

5

6

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46

Pri

ce

Ch

an

ge

(£

/M

Wh

)

Settlement Period

Graph 28A– Average short difference in System Prices

0

5

10

15

20

25

30

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46Syste

m P

rice

(£

/M

Wh

)

Settlement Period

Live Price Nov-18 Price

Graph 26 - Average long System Prices

0

20

40

60

80

100

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46

Syste

m P

rice

(£

/M

Wh

)

Settlement Period

Live Price Nov-18 Price

Graph 27 - Average short System Prices

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46

Pri

ce

Ch

an

ge

(£

/M

Wh

)

Settlement Period

Graph 29A– Average long difference in System Prices


Post Implementation

Review


Graph 28 shows the magnitude of price changes, excluding those Settlement Periods where there were no changes (58% of all Settlement Periods).

0

500

1000

1500

2000

2500

-20+

-20 -

-15

-15 -

-10

-10 -

-5

-5 -

- 0

0 -

5

5 -

10

10 -

15

15 -

20

20+

Fre

qu

en

cy o

f P

rice

Ch

an

ge

System Price changes (£/MWh)

Long Short

Graph 28 – November 2018 scenario price changes


Post Implementation

Review


Case study: the biggest price difference between the live and November 2018 scenarios when the system was long

In this case study, we compare the price stack under the current pricing arrangements with those under the

November 2018 arrangements.

With a long market, the System Price is set by Sell actions (including Bids in the Balancing Mechanism).

However, we do not include all actions in the price calculation; there are a number of tagging processes to

remove some of actions from the volume of actions that are used to set the price (the “Priced Volume”):

● NIV tagging nets off the Buy and Sell actions to determine the overall direction of the system – the

Net Imbalance Volume (NIV). The smaller stack of actions is netted off the larger stack of balancing

actions, starting with the most expensive actions. Actions removed in this way are referred to as NIV

Tagged.

● The energy imbalance price is calculated based on the volume-weighted average of a defined volume

of the most expensive actions remaining. This defined volume is the Price Average Reference

Volume (PAR) and is currently 50 MWh. The PAR is due to decrease to 1MWh in November 2018.

The less expensive volumes left after NIV tagging are ‘PAR-tagged’ and not included in the priced

volume.

The largest price difference (when the market was long) between the live pricing scenario and the November

2018 scenario occurred on 12 January 2016, Settlement Period 13. The live price was -£7.97/MWh, but under

the November 2018 scenario, the price would have been -£60.75/MWh.

This price difference was because of the impact of reduced the PAR volume. Under the live scenario, the price

was set by a volume weighted average of Sell actions from four BMUs10, whereas with a PAR of 1MWh, this

meant that an average of only one action, priced at -£60.75/MWh, set the System Price. The relatively small NIV

may have also contributed to low price, as the more expensive action of -£60.75/MWh was not NIV tagged –

had there been more Buy actions taken in this half hour, this action may have been excluded.

10 Note actions from the same BMUs and at the same price have been combined.


Post Implementation

Review


P305 and price volatility

The removal of the dual cash-out price, in

combination with the changes to PAR making

prices ‘more marginal’, may have impacted the

price volatility faced by Parties.

The single System Price reflects the cost of the

underlying sell actions when the system is long,

and buy actions when the system is short.

Because of the difference between the prices of

buy and sell actions, there can be a material price

change when the system changes direction. On

average the price impact of the NIV ‘flipping’

direction is around £20/MWh.

An example of this is illustrated by Graph 29,

which shows the impact on prices when the

system direction changed from long to short on

21 April 2016. The system was only short for

three Settlement Periods on that day – 35, 36 and

37 – and the price was £108/MWh in each of

these.

Before P305, the Market Index Price was paid to

or by Parties with imbalances in the opposite

direction to the system. The Market Index Price

acted as a ‘buffer’ when the system changed

direction but the Parties’ imbalance hadn’t.

This can be illustrated by looking at the standard

deviation of prices. Graph 30 shows the monthly

standard deviation of prices that would have been

faced by a Party who was always short (“Short Party Price”), compared to the standard deviation of prices

when the system was short (“Short System

Price”).

Graph 31 shows the standard deviation of prices

that would have been faced by a Party that was

always long (“Long Party Price”), compared to

the standard deviation of prices when the system

was long (“Long System Price”). This shows

the increased price volatility in prices paid to

Parties with long imbalances following the

implementation of P305.

-1000

-800

-600

-400

-200

0

200

400

0

20

40

60

80

100

120

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46

NIV

(M

Wh

)

Syste

m P

rice

(£

/M

Wh

)

NIV (MWh) System Price

Graph 29 – System Price plotted against the NIV for 21 April 2016

Graph 30 – Standard deviation of System Prices for a Short system and for a Party that is Short (Short Party Price)

Graph 31 - Standard deviation of System Prices for a Long system and for a Party that is Long (Long Party Price)

0

10

20

30

40

50

60

70Ja

n

Mar

May

Jul

Sep

Nov

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

2014 2015 2016

Sta

nd

ard

De

via

tio

n (

£/M

Wh

)

Short System Price Short Party Price (SBP)

0

5

10

15

20

25

30

35

40

45

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

May

Jul

Sep

Nov

Jan

Mar

2014 2015 2016

Sta

nd

ard

De

via

tio

n (

£/M

Wh

)

Long System Price Long Party Price (SSP)


Post Implementation

Review


4. IMPACT ON PARTIES TRADING CHANGES

This section provides impact analysis on BSC Parties’ Trading Charges (specifically, Imbalance Charges and

Residual Cashflow Reallocation Cashflow) by comparing live prices with the P217A and November 2018 pricing

scenarios. Note that this analysis does not take into account of behavioural changes.

The impacts on Parties are presented in daily £/MWh, which is derived from changes in £ charges divided by

the absolute Credited Energy Volume traded by Parties.

The impact analysis is broken down by BSC Party types (based on our best knowledge, see appendix two for

further information), including vertically integrated players, independent generators, renewable generators,

independent suppliers and non-physical traders. This allows us to see how different types of participants are

impacted under different pricing scenarios.

P217 Price scenario

This analysis compares Parties’ charges using live prices against P217A scenario prices (i.e. prices had P305

not been introduced).

We consider the impact on two elements of Parties’ Trading Charges that will be directly impacted by a change

to the System Price calculation:

● Daily Account Energy Imbalance Cashflow – this is the net of Imbalance Charges made by/to

Parties for imbalance volumes at the System Price; and

● Residual Cashflow Reallocation Cashflow (RCRC) – Any excess or shortfall in cashflow after

all BSC Parties have paid their Imbalance Charges is redistributed amongst BSC Parties on a

Credited Energy volume basis11. Typically, RCRC is paid to Parties, although it can also be a charge.

We have not considered the impact of other elements of Trading Charges which would not have been directly

impacted by the P305 changes – eg the BM Unit Cashflow, Non-Delivery Charges, System Operator BM

Cashflow and Information Imbalance Charges.

The average daily impacts in £/MWh are shown in Table 5 below. Positive impacts represent Parties being

‘worse off’ and negative impacts represent Parties being ‘better off’ as result of P305 when compared against

P217A scenario. The impact on Imbalance Charges should be considered alongside the impact on RCRC as it is

the net of these two which represent total charges related to imbalance.

11 Credited Energy Volume is the allocation of metered volume from BM Units to Energy Accounts in a Settlement Period,

taking account of Transmission Loss Multipliers and applying any Metered Volume Reallocation Notices that are in force

Table 5 - Average impact of P305 changes on Parties’ Trading Charges

Type of Party

Mean Imbalance

charge impact

(£/MWh)

StdDev of Imbalance

charge impact

(£/MWh)

Mean RCRC impact

(£/MWh)

StdDev RCRC impact

(£/MWh)

Independent Generator 0.11 0.80 -0.04 0.10

Non Physical Trader 0.38 2.62 0.00 0.02

Renewable Generator 4.98 8.53 -0.04 0.12

Supplier 1.45 4.32 -0.08 0.10

Vertically Integrated 0.49 3.54 -0.03 0.09


Post Implementation

Review


The majority of Parties have incurred additional Imbalance Charges following the implementation of P305.

However the magnitude of the average increase in Imbalance Charges was £2/MWh for the majority of Parties,

except for renewable generators.

Standard deviation is a measure how scattered (or varied) the data can be away from its mean value in both

directions. The standard deviation of Imbalance Charge impact on renewables generators appeared to the

highest amongst all, suggesting more variation in Imbalance Charges (in both directions) for these Parties.

This is followed by Suppliers, whose charges are more uncertain (in part due to the nature of non-half hourly

metered customer consumption). In contrast, the Imbalance Charge impacts on independent generators are of

less variance.

Residual cash-flow reallocation cashflow (RCRC) payments also increased by £0.10/MWh on a monthly average

basis.

Graph 36 shows the means and standard deviations of Imbalance Charge and RCRC impacts (in £/MWh) for

different types of Parties by month. Parties did not incur more Imbalance Charges in the winter months (i.e.

December to February), which may have been due to the mild winter in 2015/6. However there has been an

increase in charges in March and April 2016, likely to be due to the prolonged ‘winter’ and unusually low

temperature.

Graph 32 – Difference in Imbalance and RCRC Charges between P305 pricing and P217

-0.20

1.80

3.80

5.80

7.80

9.80

11.80

13.80

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hys

ical Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted


2015 2016

Pri

ce

Ch

an

ge

(£

/M

Wh

)

Mean Imbalance charge impact (£/MWh) Mean RCRC impact (£/MWh)

StdDev of Imbalance charge impact (£/MWh) StdDev RCRC impact (£/MWh)


Post Implementation

Review


Impacts on renewable generators

Graph 33 shows the trading charges impact for renewable generators on a monthly average basis.

Graph 33 – Renewable generators imbalance charge impact on changing pricing arrangements

Out of the 13 renewable generators included in this analysis, one Party had very poor balancing performance

due to ‘spilling’ with no contracts in place; P305 has reduced the imbalance payment they have received by up

to £28/MWh in April. Seven other renewable generators have been impacted by around £7/MWh on average

throughout the analysis period, whereas the remaining Parties were impacted less, with impacts ranging from

£0 to £7/MWh.

-0.1

4.9

9.9

14.9

19.9

24.9

29.9

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

Nov

Jan

Mar

2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016

Party A Party B Party C Party D Party E Party F Party G Party H Party I Party J Party K Party L Party M

Imp

act

(£/M

Wh

)

Average imbalance charge impact Average RCRC impact


Post Implementation

Review


November 2018 price scenario

This analysis compares Parties’ charges using live prices against the November 2018 price scenario.

The numerical statistics of impacts (£/MWh) can be seen in the table below. Note that positive impacts

represent ‘worse off’ and negative impacts represents ‘better off’ as result had the full package of P305 been

implemented when compared against the live scenario12.

The November 2018 pricing scenario resulted in additional Imbalance Charges for BSC Parties. We have seen a

similar trend to the P217A scenario analysis with an impact on most Parties of around £2/MWh per day, with

the exception of renewable generators. Standard deviations of Imbalance Charge impacts have shown a similar

trend as per P217A analysis in that renewable generators and Suppliers had more varied impacts on Imbalance

Charges.

PAR1 would have increased the total imbalance cashflow as well as Total Residual Cashflow across the

industry, and therefore, on average, Parties would have received more RCRC payment. The net impact on

RCRC would have been limited to £0.09/MWh during the analysis period.

Graph 34 displays the means and standard deviations of Imbalance Charge and RCRC impacts (in £/MWh) for

different types of Parties, during different time periods.

The average Imbalance Charge impact on independent generators appeared to be unexpectedly high in April

2016, with a high standard deviation. This was due to a new entrant Party having proportionally large

Imbalance Volumes (i.e. ‘spilling’ with no contract), which drove up the average figures. The impact analysis

suggests that this Party is ‘worse off’ in the winter 2018 scenario comparing to the live scenario, due to less

imbalance cashflow received.

12 For avoidance of doubt, these statistics reflect further charge increase/decrease on top of the live charges, not P217A charges.

Therefore this is not a direct comparison to P217A charges, which is not within the scope of this analysis.

Type of PartyMean imbalance

charge impact

StdDev imbalance

impactMean RCRC impact StdDev RCRC impact

Independent Generator 0.46 -0.04 4.28 0.10

Non Physical Trader 0.36 0.00 3.13 0.02

Renewable Generator 5.19 -0.03 9.94 0.11

Supplier 1.48 -0.07 4.83 0.10

Vertically Integrated 0.50 -0.02 3.70 0.10

Table 6 – Average impact on Trading Charges with changing the Pricing Arrangements to the November 2018 scenario


Post Implementation

Review


Graph 34 - Difference in Imbalance and RCRC Charges between P305 November 2018 pricing and live prices

-0.20

1.80

3.80

5.80

7.80

9.80

11.80

13.80

15.80

17.80

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted

Independent

Genera

tor

Non P

hysi

cal Tra

der

Renew

able

Genera

tor

Supplie

r

Vert

ically

Inte

gra

ted


2015 2015 2016 2016 2016 2016

Pri

ce

Ch

an

ge

(£

/M

Wh

)

Mean imbalance impact Mean RCRC impact StdDev imbalance impact StdDev RCRC impact


Post Implementation

Review


5. PARAMETER ANALYSIS

This section considers the more detailed aspects of imbalance pricing. We consider:

● The impact of moving to a smaller Price Average Reference (PAR), on the number actions setting

the price, in the pre-P305, live and November 2018 Scenarios

● Incidents of the Reserve Scarcity Price during the period, and how these relate to the

Utilisation Prices of Short Term Operating Reserve (STOR) balancing capacity

● De-Rated Margins (DRMs) across the period and National Grid’s forecasts of these

● The impact of including non-BM STOR volumes in the System Price calculation

● The impact of the changed methodology for the Buy Price Adjuster

● The impact of the reduced volume of actions setting the Replacement Price (RPAR)

Reduction in the Price Average Reference (PAR) Value

The Price Average Reference (PAR) volume is used to tag balancing actions such that a maximum volume of

PAR MWh is used to set the Energy Imbalance Price. The value of PAR P305 introduced was 50MWh, from

500MWh previously.

The P305 workgroup noted concerns that a smaller PAR value could amplify any errors or inefficiencies in the

current calculation, as there would be a smaller number of actions setting the price. Potential errors or

inefficiencies included:

● Incorrect flagging of the prices of system balancing actions by the Transmission Company; or

● The impact of plant dynamics, leading to a high-priced Offer being accepted in one Settlement

Period to resolve an issue at that time, because of the BMU’s physical abilities, the Offer may have

to persist for a longer period, impacting future Settlement Periods where a lower-priced Offer

would otherwise have been accepted.

Table 7 shows how many

balancing actions were left in the

PAR in the assessment period and

how this compares to the pre-305

PAR (500MWh) and the

November 2018 PAR (1MWh)13.

This shows that the average

number of actions in the PAR

decreases as the PAR value

decreases. It also shows that the

number of Settlement Periods

with only one action setting the price increases as the PAR value decreases. In the PAR50 and PAR1 scenarios,

the most frequent number of actions setting the price was 1 (although the frequency was greater in the PAR1

scenario).

The potential for erroneous system flags was mentioned as a concern during the P305 workgroup. National

Grid’s system flagging process is the ex-ante identification of BMUs or individual actions which could potentially

13 Separate actions from the same BMUs within a Settlement Period were treated as a single action for the purposes of this analysis.

Table 7 - Average number of balancing actions left in PAR

Average

number of

actions in the

PAR

Most frequent

number

of actions in the

PAR

Max number of

actions

in the PAR

Proportion of

SP's with only

one action in

the PAR

PAR500

(P217a) 5.19 2.00 23.00

13%

PAR50 (live) 2.40 1.00 23.00 37%

PAR1 (Nov-18) 1.63 1.00 23.00 69%


Post Implementation

Review


be classified as system balancing actions (eg actions that were taken to

resolve locational constraints). The process for this is set out in their

System Flagging Methodology (SMAF) document14.

If the Transmission Company or another Party identifies an action which

should have been SO-flagged (or erroneously received an SO flag) these

can be corrected using the BSCP18 process. This must be brought to the

attention of the Transmission Company a minimum of eight Working Days

before the SF Settlement Run. Table 8 shows the number of corrections

to a System Flag since the implementation of P305.

Reduction in the Replacement Price Average Reference (RPAR)

The Replacement Price is the price used for any actions that have become un-priced due to flagging and

classification. Flagging is the identification of actions which may have been taken for ‘system’ management

reasons (e.g. locational constraints). If a flagged action is out-of-merit (i.e. an action that is more expensive

than the next most expensive flagged action) it will lose its price, and requires a Replacement Price.

The Replacement Price is set in a similar way as the final Imbalance Price – but only actions with their original

price are used, and the weighted average used to set the price is a smaller volume. Before P305 this volume

(the Replacement Price Average Reference, RPAR) was the most expensive priced 100MWh (“RPAR100”). P305

reduced the RPAR volume to 1MWh (“RPAR1”).

Graph 39 shows average Replacement

Prices across the day, split by a long and

short system, and how the reduced RPAR

volume has affected them. As expected,

Replacement Prices are higher when the

system is short. When the system is short,

the Replacement Price tends to re-price

flagged actions downwards – actions were

priced by an average of £54.66/MWh under

PAR100, and £45.40/MWh under PAR1.

When the system was long, the

replacement price was lower under RPAR1

than RPAR100. Under RPAR100, the initial

price of actions was re-priced upwards by

an average of £31.06/MWh and under

RPAR1 actions were re-priced upwards by

an average of £21.56/MWh.

14 This is one of the Transmission licence C16 statements, and can be found at: http://www2.nationalgrid.com/uk/industry-information/electricity-codes/balancing-framework/transmission-

license-c16-statements/

Table 8 – Number of flags resubmitted by National Grid

MonthIncidents of flags

resubmitted

November 26

December 12

January 8

February 3

March 5

April 0

Total 59

0

20

40

60

80

100

120

140

160

1 3 5 7 9

11

13

15

17

19

21

23

25

27

29

31

33

35

37

39

41

43

45

47Re

pla

ce

me

nt

Pir

ce

(£

/M

Wh

)

RPAR1 - Long RPAR1 - Short

RPAR100 - Long RPAR100 - Short

Graph 35 - The impact of changing RPAR on Replacement Prices

http://www2.nationalgrid.com/uk/industry-information/electricity-codes/balancing-framework/transmission-license-c16-statements/

http://www2.nationalgrid.com/uk/industry-information/electricity-codes/balancing-framework/transmission-license-c16-statements/


Post Implementation

Review


De-Rated Margin (DRM), Loss of Load Probability (LoLP) and the Reserve Scarcity Price

P305 introduced the Reserve Scarcity Price (RSP)

which uplifts the prices of STOR balancing

capacity when it is higher than the capacity’s

original Utilisation Price.

The RSP is designed to respond to capacity

margins so that it rises as the system gets tighter

(ie the gap between available and required

generation narrows). It is a function of De-Rated

Margin (DRM) at Gate Closure, the likelihood that

this will be insufficient to meet demand (the Loss

of Load Probability, LoLP) and the Value of Lost

Load (VoLL, currently set at £3,000/MWh). For each DRM, there is an associated LOLP, which is multiplied by

VoLL to determine the RSP15.

Graph 36 shows the daily minimum and average De-rated margin since data was available (7 March 2016).

Table 9 shows the LoLPs and RSPs that the 10 lowest DRMs equated to16. The System was long in six of these

ten Settlement Periods, with all System Prices in these Settlement Periods less than £24.10/MWh.

15

The System Operator has determined a relationship between each DRM and the LoLP which will determine the RSP. The methodology

for LOLP is set out in the LOLP Methodology statement: https://www.elexon.co.uk/wp-content/uploads/2014/10/37_244_11A_LOLP_Calculation_Statement_PUBLIC.pdf 16 Due to data issues (see appendix 3) DRM, LOLP and RSP data was only operational from Settlement Period 27 on 7 March 2016.

Date SPDRM

(MW)LoLP RSP RSP Used

System

Length

System Price

(£/MWh)

10/03/2016 37 1,496.48 0.0163 48.88 No Short 377.55

10/03/2016 38 1,582.23 0.0119 35.73 No Short 517.55

19/04/2016 36 1,800.67 0.0050 15.13 No Long 24.10

07/03/2016 37 1,813.48 0.0048 14.40 No Short 135.00

08/04/2016 19 1,832.39 0.0044 13.30 No Long 24.03

15/04/2016 18 1,855.79 0.0040 12.02 No Short 131.80

19/04/2016 37 1,864.40 0.0039 11.62 No Long 24.10

20/03/2016 39 1,881.98 0.0036 10.76 No Short 151.38

19/04/2016 38 1,887.13 0.0035 10.54 No Long 24.56

07/03/2016 39 1,965.74 0.0025 7.46 No Short 98.42

Graph 36 - Daily Minimum and average DRMs

0

2000

4000

6000

8000

10000

12000

14000

16000

07-M

ar

10-M

ar

13-M

ar

16-M

ar

19-M

ar

22-M

ar

25-M

ar

28-M

ar

31-M

ar

03-A

pr

06-A

pr

09-A

pr

12-A

pr

15-A

pr

18-A

pr

21-A

pr

24-A

pr

27-A

pr

30-A

pr

MW

Min of De-rated Margin Average of De-rated Margin

Table 9 – Ten lowest De-Rated Margins and corresponding Reserve Scarcity Prices




Post Implementation

Review


The RSP will then be used to re-price STOR actions in the Imbalance Price calculation if it is higher than the

original Utilisation Price of the STOR capacity. No STOR actions were re-priced using the RSP in 2015/16, as

RSPs did not exceed STOR Utilisation Prices in any periods.

Error! Reference source not found. shows the ‘merit order’ of STOR Utilisation Prices from 7 March 2016 to 30 April 2016. In order for the RSP to be used in the imbalance price calculation, it would need to be higher than

the Utilisation Price of STOR capacity used in that same Settlement Period. The highest RSP of £48.88/MWh shown (set by a DRM of 1496.48MW), but the lowest STOR Utilisation Price observed was £63.92/MWh. Table

10 shows STOR Utilisation Price data split into price bands.

Graph 37 Utilisation Prices of STOR plant, split by BM and non-BM capacity

Table 10 - Utilisation Prices of STOR plant

48.88

0

100

200

300

400

500

600

0 2000 4000 6000 8000 10000 12000 14000

Uti

lisa

tio

n P

rice

(£

/M

Wh

)

STOR Actions Taken

BM Non-BM Highest RSP

Price (£/MWh) BM STOR Non-BM STOR

60-80 0.00% 33.72%

80-100 0.00% 56.59%

100-120 0.18% 4.45%

120-140 11.49% 2.30%

140-160 67.59% 1.12%

160-180 7.97% 0.47%

180-200 11.15% 0.00%

200-220 0.00% 0.08%

240-260 0.00% 0.95%

280-300 0.00% 0.07%

300-320 0.00% 0.26%

500-520 0.36% 0.00%

540-560 1.25% 0.00%


Post Implementation

Review


DRM as an informational tool National Grid’s one hour forecast of DRM is the value used to set the LoLP and therefore the RSP. Indicative

forecasts of DRM and LoLP are also available from eight, four and two hours ahead of the start of each

Settlement Period, and values are also calculated at 12:00 each day for all Settlement Periods up to the end of

the next Operational Day (defined under the Grid Code as the period from 05:00 on one day to 05:00 on the

following day) for which Gate Closure has not yet passed.

These forecasts may act as an additional

informational tool for market participants

to drive balancing and prices in the

Balancing Mechanism, particularly when

margins are low. With the limited data

available, we considered whether DRMs

appear to have had an impact on System

Prices.

Graph 38 shows DRM plotted against

System Price for a short System. A weak

correlation can be seen. Prices become

higher as the DRM becomes lower but

there are many other factors affecting the

price (e.g. the length of the NIV and the

time of day).

We also considered how the forecasts of

the five lowest DRMs evolved across the

different forecasts, see Graph 39. In the

case of the five lowest 1 Hour Ahead DRM

submissions, the difference from the 8

Hour Ahead submissions can appear quite

substantial. This is specifically the case for

the two lowest 1 Hour Ahead DRMs (SP 37

and SP38 on 10 March) which were below

500MWh at the 8 Hour Ahead submissions

and ended up circa 1,500MW by the time

of the 1 Hour Ahead submissions. Note

that the System Price for these two

Settlement Periods reached £377.55/MWh

and £517.55/MWh for SP37 and SP38

respectively.

Graph 38 – DRM plotted against System Price, where the System is short, with a logarithmic trend line added

0

100

200

300

400

500

600

0 5000 10000 15000 20000 25000

Syst

em

Pri

ce (

£/M

Wh

)

DRM

Log. (System Price)

Graph 39 - Five lowest DRM forecast variations throughout each submission

0

500

1000

1500

2000

2500

3000

DRM 8 HourAhead

DRM 4 HourAhead

DRM 2 HourAhead

DRM 1 HourAhead

De

-Ra

ted

Ma

rgin

(M

W)

10Mar-16 (SP 37) 10Mar-16 (SP 38)19Apr-16 (SP 36) 08Apr-16 (SP 19)15Apr-16 (SP 18)


Post Implementation

Review


Graph 40 shows the difference in DRM values between the indicative submissions (8, 4 and 2 Hour Ahead)

and the final 1 Hour Ahead submission, which will set the LoLP and the RSP.

The 8 Hour Ahead is, on average, the least accurate (an average of 869.07MW from the final value), whilst the

2 Hour Ahead is, on average, the closest to the final 1 Hour Ahead submission (an average of 240.85MW from

the final value). The graph also indicates that when the final DRM submissions are at their lowest levels

(<2000MW), the variation between the 1 hour ahead submissions and the 8 Hour Ahead submissions were

less than average (an average difference of 787.97MW).

Most variation in values came when final DRMs were between 5000MW and 10000MW. Most final DRM

submissions (43.2%) fell within this range.

Graph 40 - Correlation between accuracy of submissions and DRM values

-4000

-3000

-2000

-1000

0

1000

2000

3000

4000

5000

6000

0 5000 10000 15000 20000 25000

Dif

fere

nce

be

twe

en

su

bm

issio

ns a

nd

1

Ho

ur

Ah

ea

d D

RM

(M

W)

1 Hour Ahead DRM submissions (MW)

8 Hour Ahead Vs. 1 Hour Ahead 4 Hour Ahead Vs. 1 Hour Ahead

2 Hour Ahead Vs. 1 Hour Ahead


Post Implementation

Review


The impact of including non-BM STOR volumes in the System Price calculation

Non-Balancing Mechanism Short Term

Operating Reserve (Non-BM STOR) is

reserve capacity that is not dispatched in

the Balancing Mechanism (BM). It can be

used by the SO to balance the system

during STOR Availability Windows. P305

introduced changes to include volumes and

prices for non-BM STOR actions in the stack

of balancing actions used to determine the

imbalance price. In addition to this, non-BM

STOR may be re-priced using the Reserve

Scarcity Price (RSP). Just under 3% of total buy balancing actions taken during the assessment period were

STOR actions. Of these, 70% came from

non-BM STOR plant (see Graph 41).

To assess the impact of including these

non-BM STOR volumes we re-ran the price

calculation without non-BM volumes.

Including non-BM STOR volumes had a

marginal impact on average System Prices

– prices were unchanged in 98% of long

Settlement Periods and 93% of short

Settlement Periods. However, there were

incidents of more material impact – the

inclusion of non-BM STOR volumes caused

the system to ‘flip’ from being net long to

net short in 24 Settlement Periods across

the period.

Graph 42 shows the price changes of

including non-BM in the price stack (when

there was a price change).

The price impact of including non-BM STOR

can also be seen in Graph 43 which shows

average System Prices by Settlement Period

when non-BM STOR was utilised.

Graph 42 - Price changes - non-BM STOR

0

10

20

30

40

50

60

<1

1 -

2

2 -

3

4 -

5

6 -

7

7 -

8

8 -

9

9 -

10

10 -

15

15 -

20

20 -

30

30 -

40

40 -

50

50 -

90

Co

un

t o

f S

ett

lem

en

t P

eri

od

s

Change in price (£/MWh)

Long Short

Graph 41 – BM STOR and no-BM STOR volumes used

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

Nov

Dec

Jan

Feb

Mar

Apr

2015 2016V

olu

me

of

acti

on

s (

MW

h)

BM STOR Non-BM STOR

Graph 43 – Price impact of including non-BM STOR in the price stack, in Settlement Periods when non-BM STOR was used

0

20

40

60

80

100

120

140

160

180

1 2 3 4 5 6 7 8 9 10111213141516171819202122232425

Pri

ce

(£

/M

Wh

)

Live System Prices - Long No NBM STOR - Long

Live System Prices - Short No NBM STOR - Short


Post Implementation

Review


The Buy Price Adjuster

The Buy Price Adjuster (BPA) is an adjustment made to the System Price to reflect of the long-term

contracts that the SO enters into to provide balancing services. The BPA is added to the System Price when the

net imbalance of the Transmission System is short.

Before P305 was implemented, STOR availability costs were allocated to Settlement Periods using the BPA

according to a historic weighted profile; this approach meant that BPAs didn’t necessarily align with tight

margins or STOR usage. P305 introduced the Reserve Scarcity Price as a mechanism for targeting the value of

STOR usage, and removed STOR availability costs from the BPA calculation.

National Grid has been recalculating the STOR availability costs that would have been included in the BPA for

use within ELEXON’s analysis. Using this, we can create a ‘P217A BPA’ to see the impact that this proportion

has had on BPAs and System Prices.

Overall, the BPA is used less frequently – the live BPA applied to 6% of short Settlement Periods, whereas

P217A BPAs would have applied 46% of the time.

Because the BPA is no longer being used to recover STOR availability costs, BPAs are now lower. The live BPA

averages £4.97. Applying the P217A BPA price over the entire period (November 2015 – April 2016) would

have resulted in an average BPA of £5.69

The highest live BPA was £21.29 on the 8 January 2016, under P217A prices this would have been £49.25. The

median BPA under both P217A and the current scenario is £0, since the adjuster is only applied in a minority of

short Settlement Periods. The median BPA, when it is applied is £3.33 and £4.22 under P217A.


Post Implementation

Review


LIST OF APPENDICES

1. Glossary

2. Classification of Party types

3. Summary of data issues


Post Implementation

Review


APPENDIX ONE - GLOSSARY

Term Abbrev. Definition

Bid A proposed volume band and price within which the registrant of a BM Unit is willing to

reduce generation or increase consumption (i.e. a rate below their FPN).

Bid/Offer

Acceptance

BOA A Bid or Offer within a given Settlement Period that was Accepted by the SO. BOAs are

used in the imbalance price calculation process e.g. to calculate NIV or the System Price.

Offer A proposed volume band and price within which the registrant of a BM Unit is willing to

increase generation or reduce consumption (i.e. a rate above their FPN).

System Price A price (in £/MWh) calculated by BSC Central Systems that is applied to imbalance

volumes of BSC Parties. It is a core component of the balancing and settlement of electricity in GB and is calculated for every Settlement Period. It is subject to change via

Standard Settlement Runs.

Replacement Price

A price (in £/MWh) calculated by BSC Central Systems that is applied to volumes that are not priced during the imbalance pricing process (detailed in BSC Section T) It is calculated

for every Settlement Period, and is subject to change via Standard Settlement Runs.

Utilisation Price The price (in £/MWh) sent by the SO in respect of the utilisation of a STOR Action which: (i) in relation to a BM STOR Action shall be the Offer Price; and

(ii) in relation to a Non-BM STOR Action shall be the Balancing Services Adjustment Cost.

Market Price The Market Price reflects the price of wholesale electricity in the short-term market (in £/MWh). You can find an explanation of how it is calculated and used in the Market Index

Definition Statement (MIDS).

Reserve Scarcity Price

RSP Both accepted BM and non-BM STOR Actions are included in the calculation of System Prices as individual actions, with a price which is the greater of the Utilisation Price for

that action or the RSP. The RSP function is based on the prevailing system scarcity, and is

calculated as the product of two following values: Loss of Lost Load (LoLP), which will be calculated by the SO at Gate Closure for

each Settlement Period; and currently set to £3,000/MWh.

Replacement

Price Average Reference

RPAR The RPAR volume is a set volume of the most expensive priced actions remaining at the

end of the System Price calculation, and is currently 1MWh. The volume-weighted average of these actions, known as the Replacement Price, is used to provide a price for any

remaining unpriced actions prior to PAR Tagging.

Long In reference to market length, this means that the volume of Accepted Bids exceeds that

of Accepted Offers

Short In reference to market length, this means that the volume of Accepted Offers exceeds

that of Accepted Bid

Net Imbalance NIV The imbalance volume (in MWh) of the total system for a given Settlement Period. It is


Post Implementation

Review


Volume derived by netting Buy and Sell Actions in the Balancing Mechanism. Where NIV is positive, this means that the system is short and would normally result in the SO

accepting Offers to increase generation/decrease consumption. Where NIV is negative,

the system is long and the SO would normally accept Bids to reduce generation/increase consumption. It is subject to change via Standard Settlement Runs.


Post Implementation

Review


APPENDIX TWO - CLASSIFICATION OF PARTY TYPES

Party Group name Type of Party


Addito Supply Limited Supplier

Gold Energy Supply Ltd Supplier

AVRO ENERGY LIMITED Supplier

Good Energy Limited Supplier

Axis Telecom Ltd Supplier

Grangemouth CHP Ltd

Independent Generator

AXPO Supplier

Greater Gabbard Offshore Winds Renewable Generator

AXPO UK LIMITED Supplier

Green Energy (UK) Plc Supplier

Baglan Operations Ltd Independent Generator

Green Network Energy Limited Supplier

Barbican Power Limited Supplier

Gwynt y Mor Offshore Wind Farm Renewable Generator

Barclays Bank plc Non Physical Trader

Hanbury Energy Supplier

BARKING POWER LIMITED Independent Generator

Haven Power Ltd Supplier

Bayswater Energy Limited Supplier

Holborn Energy Limited Supplier

BES Commercial Electricity Ltd Supplier

Hudson Energy Supply UK Ltd Supplier

Bethnal Energy Limited Supplier

I Supply Energy Limited Supplier

BKW Energie AG Non Physical Trader

Iberdrola (ScottishPower) Vertically Integrated

BNP Paribas Non Physical Trader

Inovyn Chlor Energy LTD Supplier

Bord Gais Energy Limited Non Physical Trader

Intergen


BP Gas Marketing Limited Non Physical Trader

IPM Energy Retail Ltd Vertically Integrated

Breeze Energy Supply Ltd Supplier

IPM Energy Trading Limited Vertically Integrated

British Energy Direct Ltd Supplier

Iresa Limited Supplier

Bronze Energy Supply Ltd Supplier

J.Aron & Company Non Physical Trader

Brookfield Renewable Supply 2 Non Physical Trader

JP Morgan Chase Bank Non Physical Trader

BTG Pactual Commodities Non Physical Trader

Kensington Power Limited Supplier

Business Power & Gas Limited Supplier

KOCH SUPPLY & TRADING SARL Non Physical Trader

Captured Carbon Non Physical Trader

Lavender Energy Supply Ltd Supplier

Cargill PLC Non Physical Trader

Limejump Energy Limited Supplier

Carrington Power Ltd Independent Generator

Lincs Wind Farm Ltd Renewable Generator

Cenergise Limited Non Physical Trader

LOCO2 Energy Supply Limited Supplier

Centrica Vertically Integrated

London Array Limited Renewable Generator

Citigroup Global Markets Ltd Non Physical Trader

MA Energy Limited Supplier

Compagnie Nationale du Rhone Non Physical Trader

Macquarie Bank Limited Non Physical Trader

Co-operative Energy Limited Supplier

Marble Power Limited Supplier

Copper Energy Supply Ltd Supplier

Marigold Energy Supply Ltd Supplier

CORBY POWER LIMITED Renewable Generator

Mercuria Energy Trading SA Non Physical Trader

Cornflower Energy Supply Ltd Supplier

Mercury Energy Supply Ltd Supplier

Corona Energy Retail 5 Ltd Supplier

Merrill Lynch Commodities LTD Non Physical Trader

Coulomb Energy Supply Limited Supplier

Mint Energy Supply Ltd Supplier

Covent Energy Limited Supplier

Mistral Energy Supply Limited Supplier


Post Implementation

Review




Daisy Energy Supply Ltd Supplier

Morecambe Wind Limited Renewable Generator

Danske Commodities A/S Non Physical Trader

Morgan Stanley Non Physical Trader

DONG Energy Vertically Integrated

MVV Environment Services Ltd Supplier

Dong Energy Power UK Limited Vertically Integrated

NDA Supplier

Donnington Energy Limited Supplier

NEAS Energy A/S Supplier

Drax Power Ltd Independent Generator

NEAS Energy Limited Supplier

Dual Energy Direct Limited Supplier

Nickel Energy Supply Ltd Supplier

E.ON Vertically Integrated

NIE Energy Ltd Non Physical Trader

E.ON Vertically Integrated

Noble Clean Fuels Limited Non Physical Trader

ECC AG Non Physical Trader

Nord Pool Spot AS Non Physical Trader

ECONOMY ENERGY SUPPLY LTD Supplier

Northern Ireland Electricity Non Physical Trader

Economy Energy Trading Limited Supplier

Opus Energy Limited Supplier

Economy Power Supplier

Osmium Energy Supply Limited Supplier

Edelweiss Energia S.p.A Non Physical Trader

Our Power Energy Supply Ltd Supplier

EDF ENERGY Vertically Integrated

OVO Electricity Ltd Supplier

EDF Energy PLC Vertically Integrated

Paddington Power Limited Supplier

Edgware Energy Limited Supplier

Palladium Energy Supply Ltd Supplier

Effortless Energy Ltd Supplier

Petroineos Trading Ltd Non Physical Trader

Eggborough Power Limited Independent Generator

PFP Energy Supplies Limited Supplier

Electrabel SA Non Physical Trader

POWER4ALL Limited Supplier

Electricity Plus Supply Ltd Supplier

PowerQ Ltd Supplier

ElectroRoute Energy Trading Non Physical Trader

Reuben Power Supply Limited Supplier

ENDESA GENERACION SAU Non Physical Trader

Robin Hood Energy Limited Supplier

Eneco Energy Trade Supplier

RWE Npower Vertically Integrated

Enel Trade SpA Non Physical Trader

Scira Offshore Energy Limited Renewable Generator

Energi Danmark A/S Non Physical Trader

SembCorp Utilities (UK) Ltd Renewable Generator

Energy Data Company Limited Supplier

Shell Energy Europe Limited Non Physical Trader

Energy24 Limited Supplier

Silver Energy Supply Ltd Supplier

Energya VM Gestion De Energia Non Physical Trader

Sinq Power Ltd Supplier

Engie (aka GDF SUEZ) Vertically Integrated

Smartestenergy Limited Supplier

EPG Energy Limited Supplier

Snowdrop Energy Supply Ltd Supplier

Epower Supply Ltd Supplier

Spark Energy Supply Limited Supplier

Erova Energy Limited Non Physical Trader

SSE Vertically Integrated

ESB Independent Energy (NI) Non Physical Trader

Stadtwerke Munchen GmbH Renewable Generator

ESBIG Trading Ltd Non Physical Trader

Statkraft Markets Gmbh Supplier

ESSO Petroleum Company Ltd Renewable Generator

Statoil Gas Trading Ltd Renewable Generator

Extra Energy Supply Limited Supplier

Switch Business Gas and Power Supplier

F & S Energy Ltd Supplier

Symbio Energy Ltd Supplier

Fallago Rig Wind Farm Limited Renewable Generator

Tailwind Energy Supply Limited Supplier


Post Implementation

Review




Farmoor Energy Limited Supplier

Team Supplier

Farringdon Energy Limited Supplier

Tempus Energy Supply Limited Supplier

First Hydro Company Independent Generator

The Renewable Energy Co Ltd Supplier

First Utility Limited Supplier

Tornado Energy Supply Limited Supplier

Flow Energy Ltd Supplier

Total Gas & Power Ltd Supplier

Foxglove Energy Supply Ltd Supplier

Trailstone GMBH Non Physical Trader

Freepoint Commodities Europe Non Physical Trader

Tulip Energy Supply Ltd Supplier

Future Energy Utilities Ltd Supplier

Utilita Energy Limited Supplier

GAELECTRIC ICT ROI Non Physical Trader

Vattenfall Energy Trading Supplier

Gazprom M & T Retail Ltd Supplier

Vavu Power Limited Supplier

GAZPROM MARKETING & TRADING LD Supplier

VAYU LIMITED Non Physical Trader

GB Energy Supply Limited Supplier

VIRIDIAN ENERGY SUPPLY LTD Non Physical Trader

GFP Trading Ltd Renewable Generator

VITOL SA Non Physical Trader

GNERGY Limited Supplier

VPI Immingham LLP


Zeven Electricity ltd Supplier


Post Implementation

Review


APPENDIX THREE - SUMMARY OF DATA ISSUES FOLLOWING THE IMPLEMENTATION OF P305

This note provides a summary of pricing issues related to BSC Modification P305. All of these issues have now

been resolved, and for the avoidance of doubt, no retrospective price changes will be applied.

What are LOLP and DRM?

The implementation of BSC Modification 305 on 5 November 2015 created two new parameters that are used

in the calculation of imbalance prices: De-rated Margin (DRM) and Loss of Load Probability (LOLP). DRM is a

measure of available capacity and is calculated by the System Operator at gate closure. LOLP is the likelihood

that there will be insufficient capacity to meet demand. There is a LOLP for each DRM, and LOLP is a value

between 0 and 1.

DRM and LOLP may also be used in the calculation of the imbalance price. The LOLP for a given Settlement

Period is multiplied by value of lost load (VoLL, currently set to £3,000/MWh) to determine the Reserve

Scarcity Price (RSP). The RSP can then be used to re-price some actions from balancing capacity used to

provide reserve (Short Term Operating Reserve).

What was the issue – calculation and

publication of LOLP and DRM data?

1hr forecasts of DRM and LOLP are published at

Gate Closure on BMReports for every Settlement

Period, along with 12:00, 8hr ahead, 4hr ahead

and 2hr forecasts. The 1hr forecast may also be

used in the calculation of imbalance prices, if

DRMs are low enough to drive a LOLP high enough

to result in a RSP higher than the Utilisation Price

of STOR plant.

On 26 November 2015 an issue with the data was

uncovered and, following investigation, National

Grid, in consultation with ELEXON, decided to

suspend the production of DRM and LoLP from 27

November until the cause of the issue is found and

corrected.

On 8 December 2015, National Grid implemented a fix to the issue and resumed publishing DRM and LoLP

values until 27 January 2016 when a review of the data published uncovered erroneous values being produced

since 5 November 2015. National Grid, in agreement with ELEXON and Ofgem decided to stop sending LoLP

and DRM data again until the source is identified and a permanent solution developed.

Investigation and testing completed, the reporting of DRM and LoLP started again on Monday 7 March 2016

from 12:00 midday.

Following industry consultation, National Grid decided not to re-send data for the erroneous periods and

ELEXON will not recalculate System Prices retrospectively.

Since then, the BMRA experienced several smaller issues with missing DRMs and LoLPs which were re-sent by

national Grid and the last issue was resolved on 31 March 2016.


Post Implementation

Review


What was the issue with flagging of BM STOR?

In order to re-price Short Term Operation Reserve (STOR) capacity with the Reserve Scarcity Price, STOR plant

are flagged with a ‘STOR Provider Flag’. Following the implementation of P305, some STOR providers were not

being identified using the STOR Provider flag when Acceptances were reported to ELEXON.

National Grid corrected this data each day manually using a process set out in BSCP18 for amending

Acceptances after the Settlement Period in time for the SF Settlement Run or sooner. A longer term fix to this

was implemented on 7 March. There was no price impact as a result of this error.

What was the issue with reporting of non-BM STOR volumes?

Non–Balancing Mechanism STOR (Non-BM STOR) is STOR capacity that is not dispatched in the Balancing

Mechanism. P305 introduced changes to include these volumes in the stack of balancing actions used to

determine the imbalance price. However, data issues meant that these actions were reported as a single,

aggregated volume for each Settlement Period, rather than individual balancing actions. Such data issues

related to the utilisation of systems originally designed to manage the transmission system as opposed to

reporting on it.

A solution to this was implemented on 11 March 2016, and non-BM STOR volumes have since been reported

on a disaggregated basis. Overall volumes of non-BM STOR submitted do not appear to have changed as a

result, but the number of actions attributed to non-BM STOR is greater: we now receive an average of 24

individual non-BM STOR actions when it is used. The overall price impact of submitting these files on a

disaggregated basis is ambiguous, as their impact will depend on how a greater number of smaller actions

relate to other balancing actions taken in that Settlement Period.

Will prices be retrospectively changed to include the corrected data from those issues?

As mentioned in ELEXON Circular EL02398, National Grid is minded not to retrospectively apply disaggregated

volumes and re-issue this data for recalculation of System Prices. This is in response to limited industry

feedback indicating a strong direction either way. To date, ELEXON is not due to recalculate and publish

impacted System Prices retrospectively.

In any case, National Grid will provide historical disaggregated non-BM STOR utilisation volumes which will be

published on the ELEXON Portal.

National Grid acknowledges that a number of issues were experienced following the development and

deployment of the Market Operation Data Interface System (MODIS) relating to data feeds in support of BSC

modification P305 that implemented Ofgem's Electricity Balancing Significant Code Review (EBSCR). Following these issues National Grid has conducted a series of exercises to understand what went wrong, implement a

plan to correct the issues on an interim and enduring basis as appropriate, improve and embed business processes as well as ensuring learnings are transferred to future development activities. The outcome of these

exercises has been summarised in an EBSCR Project Lessons Learned Report17.

17 This report can be found at: http://www2.nationalgrid.com/UK/Industry-information/Europe/ETR-E-Modis/

https://www.elexon.co.uk/wp-content/uploads/2016/03/EL02398.pdf

http://www2.nationalgrid.com/UK/Industry-information/Europe/ETR-E-Modis/

Documents

POST IMPLEMENTATION REVIEW OF P305 - ELEXON · BSC Modification P305 introduced a number of changes to the calculation of the cash-out price, and was implemented on 5 November 2015