Defining Locational Procurement for NESO in Great Britain

Summary

In this case study, you will discover how N-SIDE and NESO co-designed a model to integrate transmission constraints and reserve locations directly into balancing capacity auctions. This approach enables System Operators to procure reserves which can be activated without causing network congestion. The study highlights:

• Market Design: Co-optimising balancing reserves with physical grid constraints.
• Product Activation: The impact of frequency-driven vs. dispatch-driven services on market clearing.
• GB Impact: Quantifiable efficiency gains and expected savings for the Great Britain system.

Introduction

N-SIDE conducted a quantitative evaluation for NESO to determine the potential social welfare benefits of introducing locational procurement into the GB system, and considering the potential cost-efficiencies for the end consumer. Previous to these findings, as the provider for GB’s Enduring Auction Capability (EAC), we developed the design modifications required to integrate transmission constraints within the clearing engine for the daily procurement of Ancillary Services.

Customer Challenge

NESO procures balancing capacity via day-ahead auctions, but grid congestion often “sterilises” these reserves, effectively trapping them behind bottlenecks where they’re unusable. To guarantee deployable reserves, NESO must replace this capacity by redispatching units via the Balancing Mechanism. This ensures system security but comes at a significant cost; projections show annual constraint costs could peak at £8 billion by 2030 without urgent market and network reform. Sterilised reserves will account for a growing share of that projected burden.

Solution/Approach

A sound design for modeling transmission constraints in balancing capacity markets should evolve around three topics:

• Transfer limits: Defining the volume transferable between zones.
• Service requirements: Identifying global and local needs per service.
• Multi-zone transfer rules: Determining if capacity in one zone can meet the requirements of another.

The answer to these questions depends fundamentally on the activation instructions of the types of reserves. In GB, balancing services are grouped into two product types: Frequency Response and Reserve. We list below the key differences:

Similar to Frequency Containment Reserves (FCR) in Continental Europe, Frequency Response is activated proportionally to frequency deviation. Because these flows cannot be controlled, N-SIDE’s model ignores transmission constraints for this product. This is similar to the FCR Cooperation, where no reservation of cross-zonal capacity is needed. 

For Reserve, network constraints must be explicitly specified via Available Transfer Capacity (ATCs) for each boundary, direction, and time window. The ATC between two adjacent zones represents the amount of transmission capacity that N-SIDE’s clearing engine identifies as available to allocate for the transfer of reserve services.

A summary of the design recommended by N-SIDE is depicted below:

Implementation

N-SIDE conducted a simulation exercise to quantify the business case for locational procurement by estimating the potential reductions in overall balancing costs for the GB market that could result from a change in market design. This study compared national procurement against locational procurement of Frequency Response and Reserve services within 5-zone and 12-zone GB network models.

The exercise adapted the existing market-clearing algorithm to consider locational features, and used 405 days of historical buy and sell orders, together with historical data on network congestion, represented as limits on the maximum quantity of reserve services cleared in certain zones.

Benefits

The modeling assumes that the 12-zone model represents all network constraints, meaning that no congestions are expected within the zone. Under this consideration, we could conclude that: 

• In the 12-zone model, locational procurement mitigated 2.2 TW of sterilised reserve capacity (including both positive and negative reserve capacity) over the modelled time period. This represented approximately 6.2% of all reserve capacity procured over the period. 
• Total auction clearing costs for reserve increased by £9.3 million or 16% over the period, as a result of selecting more expensive (but deployable) units and the consequent increase in auction clearing prices. However, if we assume that the £/MW/h cost of repositioning the 2.2 TW of sterilised capacity in the BM is equal to NESO’s auction bid price (i.e., the price of the most expensive buy order), then there are savings of £16.4 million in repositioning costs in the BM. The net benefit is therefore £7.1 million, or 11.4% of current auction procurement costs. 

We believe that our assumption regarding repositioning costs is conservative. If, in fact, the cost of BM actions to mitigate constraints is higher than the bid price of NESO’s EAC buy order for reserve, then the benefits of locational procurement would be commensurately higher.

Conclusion

The project outcomes were presented by NESO in a webinar, outlining the strategic timeline for the future of Locational Procurement in Great Britain.

For N-SIDE, the key outcomes of the project were:

• Economic Impact: Locational procurement is a proven solution for reducing the significant repositioning costs caused by “sterilised capacity.” Given that thermal constraints affect many global grids, N-SIDE believes this methodology is highly applicable to other geographies. 
• System Reality: The project remarks the necessity of market designs that reflect physical power system realities, specifically regarding product activation types and network constraint representation.
• Scalable Innovation: Successful delivery required a strong link between market design and auction clearing to ensure the algorithmic tractability of these complex market evolutions.