DCC Service Level Agreements for the Switching Incentive Regime
Summary
Requirements for the Central Switching Service (CSS) were identified during the Design, Build and Test Phase of the Ofgem-led Switching Programme as a basis for the procurement of Switching systems. These requirements were later baselined in the REC Service Definitions as Service Level Agreements (SLAs). Following the successful go-live of the CSS, these SLAs merit a review, in order to ensure that they are realistic, given the physical systems design and the actual behaviours of users, compared to the assumptions that were made in the early design phase of the Switching Programme. This Change Proposal has been raised as the vehicle to complete the review and take forward any identified changes.
Latest update and next planned activity
Date of latest update
Latest Update
22/07/2024 - The Authority rejected R0092 and approved R0092A for implementation 08/11/2024. R0092A Change Proposal page link: https://recportal.co.uk/group/guest/-/dcc-service-level-agreements-for-the-switching-incentive-regime-1
Impacts Overview
Impacted Consumers
Others
Non-Domestic
Domestic
Impacted REC Parties and Market Participants
Electricity Distribution Network Operator (R)
Data and Communications Company (DCC)
Shipper (SHP)
Energy Supplier (ES)
Gas Transporter (GT)
Impacted REC Documents
Schedule 6 - Performance Assurance
Schedule 24 - Switching Data Management
Impacted REC Services
CSS Certificate Authority
Central Switching Service (CSS)
Priority and rationale
High
Solution development
Problem Statement
Following the introduction of Performance Charges, on 01 April 2023, charges will apply based on the CRS Provider's performance against the defined SLAs. The SLAs were developed based on assumptions made in the early phases of the Switching Programme. The DCC has proposed that these are reviewed, to ensure they are fit for purpose and any Performance Charges applied are proportionate to CSS User detriment and Consumer impact.
Proposed solution summary
- As a Consumer, I want the CRS Provider to provide an effective service so that my Switch is not delayed.
- As the REC Code Manager, I want to ensure the CRS Provider has appropriate SLAs set, which incentivises the effective functioning of the Switching Service.
- As the CRS Provider, I want to ensure that the appropriate SLAs are in place to incentivise the effective functioning of the Switching Service and if these SLAs are not met by the CRS Provider, the commercial impact is proportionate to the detriment experienced by CSS Users and Consumers.
Solution documents
Documents Library
05 Jun 2026
Download R0092A_Authority_Decision (22.07.2024)
Change Plan
Solution Development
Completed
Start:
01 Feb 2023Due:
05 Oct 2023Completion:
05 Oct 2023Service Provider Impact Assessment
Completed
Start:
05 Oct 2023Due:
20 Nov 2023Completion:
13 Nov 2023Party Impact Assessment
Completed
Start:
20 Nov 2023Due:
03 Jan 2024Completion:
10 Nov 2023Preliminary Change Report issued
Completed
Start:
03 Jan 2024Due:
20 Feb 2024Completion:
16 Feb 2024Preliminary Change Report - Committee Review
Completed
Start:
20 Feb 2024Due:
27 Feb 2024Completion:
27 Feb 2024Consultation close
Completed
Start:
01 Mar 2024Due:
04 Apr 2024Completion:
04 Apr 2024Final Change Report issued
Completed
Start:
04 Apr 2024Due:
23 Apr 2024Completion:
19 Apr 2024Responsible Committee Decision
Completed
Start:
23 Apr 2024Due:
30 Apr 2024Completion:
30 Apr 2024Sent to Authority for Determination
Not Started
Start:
30 Apr 2024Due:
01 May 2024Completion:
Consultation published
Completed
Start:
27 Feb 2024Due:
01 Mar 2024Completion:
01 Mar 2024Original submission
Change details
Cordelia Grey - DCC
1 Feb 2023
Change details
DCC Service Level Agreements for the Switching Incentive Regime
Requirements for the Central Switching Service (CSS) were identified during the Design, Build and Test Phase of the Ofgem-led Switching Programme as a basis for the procurement of Switching systems. These requirements were later baselined in the REC Service Definitions as Service Level Agreements (SLAs).
Following the successful go-live of the CSS, these SLAs merit a review, in order to ensure that they are realistic, given the physical systems design and the actual behaviours of users, compared to the assumptions that were made in the early design phase of the Switching Programme.
Such a review of SLAs – with a view to maintaining requirements that provide systems performance levels required by REC Parties and ensuring that relevant performance aspects are being measured, reported on and incentivised – would provide greater clarity to REC parties and more effective performance management
of CSS performance.
Since Ofgem has decided that the SLAs from the REC Service Definitions should be used as the basis for a Switching Incentive Regime, it is also important to ensure that relevant and realistic SLAs are identified in order incentivise performance on relevant aspects of CSS performance.
Examples of where the current SLAs could be improved are:
• As customers have no requirement or ability to know whether they are interacting with CSS during times of an ‘average’ or ‘peak’ load on the CSS, their expectation should be based on a single value, aligned to the impact on their use of the service.
• For synchronous message responses, a single target should would suffice and provide clarity, irrespective of the time of day or other users’ activities occurring at the same point in time.
• Gate Closure times are currently based on arbitrary values of 20 minutes (average) or 35 minutes (peak). However, from customer feedback, detriment would not be expected to occur until at least an hour had elapsed.
• It is also key to note that many customer gateways cannot handle the existing load which results in messages not being accepted by their systems until later in the day (within a 12 hour retry period) without detriment.
Examples of options to address this include:
• A three second SLA for synchronous messages which would create a consistent baseline target (irrespective of volume)
• A five second SLA for asynchronous messages which would create a consistent baseline target (irrespective of volume)
• A one-hour window for Gate Closure which would give certainty and meet users stated requirements, but would reduce utilisation of costly resources, and balance the load to external users.
Following the successful go-live of the CSS, these SLAs merit a review, in order to ensure that they are realistic, given the physical systems design and the actual behaviours of users, compared to the assumptions that were made in the early design phase of the Switching Programme.
Such a review of SLAs – with a view to maintaining requirements that provide systems performance levels required by REC Parties and ensuring that relevant performance aspects are being measured, reported on and incentivised – would provide greater clarity to REC parties and more effective performance management
of CSS performance.
Since Ofgem has decided that the SLAs from the REC Service Definitions should be used as the basis for a Switching Incentive Regime, it is also important to ensure that relevant and realistic SLAs are identified in order incentivise performance on relevant aspects of CSS performance.
Examples of where the current SLAs could be improved are:
• As customers have no requirement or ability to know whether they are interacting with CSS during times of an ‘average’ or ‘peak’ load on the CSS, their expectation should be based on a single value, aligned to the impact on their use of the service.
• For synchronous message responses, a single target should would suffice and provide clarity, irrespective of the time of day or other users’ activities occurring at the same point in time.
• Gate Closure times are currently based on arbitrary values of 20 minutes (average) or 35 minutes (peak). However, from customer feedback, detriment would not be expected to occur until at least an hour had elapsed.
• It is also key to note that many customer gateways cannot handle the existing load which results in messages not being accepted by their systems until later in the day (within a 12 hour retry period) without detriment.
Examples of options to address this include:
• A three second SLA for synchronous messages which would create a consistent baseline target (irrespective of volume)
• A five second SLA for asynchronous messages which would create a consistent baseline target (irrespective of volume)
• A one-hour window for Gate Closure which would give certainty and meet users stated requirements, but would reduce utilisation of costly resources, and balance the load to external users.
In the long term, it will be more costly to end users to meet existing SLAs as the CSS will be required to scale and use more resources than it otherwise would if the load was balanced using less resources, over a longer period of time.
Changes to Paragraph 6.1 of the Service Levels: The REC requires a ‘mean average’ for the performance parameter of ‘average daily volume’ and ‘peak daily volume’ however, this results in the reporting of a percentage of hours where the target has not been met rather than recording the time each individual message takes which we believe would be more beneficial and would require the setting of specific targets at different time thresholds. We also propose a single target rather than different targets for average and peak which would remove the 90% percentile time performance levels.
Reporting Exclusion for Message Response Targets:
To address one of the major reasons for slow responses we propose to exclude instances where Users send large batches of messages to CSS which requires the CSS to scale up very quickly. This is not reflective of the day-to-day use of the service by REC Parties and not how the system was designed to operate.
Message Response Targets (Paragraph 6.2)
Changes to Paragraph 6.2 of the Service Levels: We propose to record the time of each individual message instead of mean averages and setting out specific targets at different time thresholds, such as target times and minimum times.
The thresholds for ‘average hourly volume’ and ‘peak hourly volume’ require a review so that the total message volumes are considered and not just the switch requests, or a single target is created rather than different targets for average and peak.
The 90th percentile and mean time of five seconds or less performance levels would no longer be applicable if a single target measure was utilised instead.
Reporting Exclusion for Message Response Targets:
Similarly to paragraph 6.1 exclusions we propose to exclude instances where Users send large batches of messages to CSS which requires the CSS to scale up very quickly. This is not reflective of the day-to-day use of the service by REC Parties and is not how the system was designed to operate.
Gate Closure Targets (Paragraph 6.3)
Changes to Paragraph 6.3 of the Service Levels: DCC propose to review the setting of thresholds for average daily volume and peak daily volume so that a single target us created instead of different targets for average and peak. Should a single target be created, this would also mean the 90th percentile measures would no longer be required.
Reporting Exclusion for Gate Closure Targets:
Similarly to paragraphs 6.1 and 6.2 exclusions, we propose to exclude instances of specific user behaviour from the reporting particularly where measures include end user system response times as these are unable to be controlled by DCC, but slow responses by users are reflected in the measure. As such, we propose to exclude user-response time from the measure, or to receive a ‘Pass’ in instances where the response time exceeds a certain value.
Incident Management Targets (Paragraph 5.2)
Changes to Paragraph 5.2 of the CSS User Support:
P1 and P2
The current target levels for all performance measures related to Incidents is set at 100%. We propose the performance measure should be set as both a target level (e.g. 100%) and a minimum service level (e.g. 85%). This measure could be set outside of the REC CSS Service Definition but within the Switching Incentive Regime itself.
Failures that occur outside of DCC’s contractual control should be excluded and referenced within the REC and to align contractually we propose that P1 and P2 measures be combined into a single ‘Major Incident SLA’.
Should the proposal above be taken forward then Priority 2 would be deleted.
P3 and P4
The current target levels for all performance measures related to Incidents is set at 100%. We propose the performance measure should be set as both a target level (e.g. 90%) and a minimum service level (e.g. 80%). This measure could be set outside of the REC CSS Service Definition but within the Switching Incentive Regime itself.
Failures that occur outside of DCC’s contractual control should be excluded and referenced within the REC and to align contractually we propose that P3 and P4 measures be combined into a single ‘Standard Incident SLA’.
Should the proposal above be taken forward then Priority 4 would be deleted.
Reporting Exclusion for Message Response Targets:
To address one of the major reasons for slow responses we propose to exclude instances where Users send large batches of messages to CSS which requires the CSS to scale up very quickly. This is not reflective of the day-to-day use of the service by REC Parties and not how the system was designed to operate.
Message Response Targets (Paragraph 6.2)
Changes to Paragraph 6.2 of the Service Levels: We propose to record the time of each individual message instead of mean averages and setting out specific targets at different time thresholds, such as target times and minimum times.
The thresholds for ‘average hourly volume’ and ‘peak hourly volume’ require a review so that the total message volumes are considered and not just the switch requests, or a single target is created rather than different targets for average and peak.
The 90th percentile and mean time of five seconds or less performance levels would no longer be applicable if a single target measure was utilised instead.
Reporting Exclusion for Message Response Targets:
Similarly to paragraph 6.1 exclusions we propose to exclude instances where Users send large batches of messages to CSS which requires the CSS to scale up very quickly. This is not reflective of the day-to-day use of the service by REC Parties and is not how the system was designed to operate.
Gate Closure Targets (Paragraph 6.3)
Changes to Paragraph 6.3 of the Service Levels: DCC propose to review the setting of thresholds for average daily volume and peak daily volume so that a single target us created instead of different targets for average and peak. Should a single target be created, this would also mean the 90th percentile measures would no longer be required.
Reporting Exclusion for Gate Closure Targets:
Similarly to paragraphs 6.1 and 6.2 exclusions, we propose to exclude instances of specific user behaviour from the reporting particularly where measures include end user system response times as these are unable to be controlled by DCC, but slow responses by users are reflected in the measure. As such, we propose to exclude user-response time from the measure, or to receive a ‘Pass’ in instances where the response time exceeds a certain value.
Incident Management Targets (Paragraph 5.2)
Changes to Paragraph 5.2 of the CSS User Support:
P1 and P2
The current target levels for all performance measures related to Incidents is set at 100%. We propose the performance measure should be set as both a target level (e.g. 100%) and a minimum service level (e.g. 85%). This measure could be set outside of the REC CSS Service Definition but within the Switching Incentive Regime itself.
Failures that occur outside of DCC’s contractual control should be excluded and referenced within the REC and to align contractually we propose that P1 and P2 measures be combined into a single ‘Major Incident SLA’.
Should the proposal above be taken forward then Priority 2 would be deleted.
P3 and P4
The current target levels for all performance measures related to Incidents is set at 100%. We propose the performance measure should be set as both a target level (e.g. 90%) and a minimum service level (e.g. 80%). This measure could be set outside of the REC CSS Service Definition but within the Switching Incentive Regime itself.
Failures that occur outside of DCC’s contractual control should be excluded and referenced within the REC and to align contractually we propose that P3 and P4 measures be combined into a single ‘Standard Incident SLA’.
Should the proposal above be taken forward then Priority 4 would be deleted.
Please see the attached slides.
Impacts
Electricity, Gas, Domestic Prepayment, Domestic Credit, Microbusiness, Small and Medium Enterprise (SME), Industrial and Commercial (I&C)
Trust and confidence
Consumers will be impacted in an indirect way, due to more effective and efficient performance management of the CSS.
No details provided
CSS Service Definition
No details provided
-
No details provided
-
No details provided
-
No details provided
-
No details provided
-
No details provided
-
No details provided
-
-
No details provided
Priority & timing
No
-
See above. These changes should be made by 1 April 2023.
A Decision from Ofgem on REC CP R0025 is expected.
Ofgem is currently working on the Direction and Guidance to give effect to a Switching Incentive Regime.
Ofgem is currently working on the Direction and Guidance to give effect to a Switching Incentive Regime.
No details provided
-