Grid charging overhaul can deliver zonal pricing benefits without ‘massive upheaval’

Scottish Power has claimed its proposed overhaul of locational transmission charges would provide much of the benefits of zonal power pricing, while avoiding the “massive upheaval” that its  introduction would also bring.

Speaking to Utility Week, senior figures at the company said its charging model would create the same long-run investment signals and also reduce the volatility and unpredictability of transmission charges – characteristics they argue would also be exhibited by zonal pricing.

Earlier this month, Scottish Power proposed a modification to the Connection Use of System Code (CUSC), designated CMP433, to implement its Optimised Transmission Investment Cost (OpTIC) charging model. This would replace the Investment Cost Related Pricing (ICRP) methodology currently used to determine wider zonal Transmission Network Use of System (TNUoS) charges.  

Joseph Dunn, head of grid and regulation at Scottish Power Renewables and the named proposer of the modification, said they started developing the model more than two years ago, prior to the launch of the government’s Review of Electricity Market Arrangements (REMA).  

Following “decades of incremental improvements”, Dunn said they wanted to find a “more enduring” solution to the issues that have plagued locational TNUoS charges, namely that they are “volatile and quite hard to predict”. 

Perhaps the most controversial option being considered by the government as part of REMA is locational marginal pricing (LMP), which would see wholesale power prices vary across a series of zones or nodes on the transmission network and reflect the balance of supply and demand in each location.  

The Department for Energy Security and Net Zero (DESNZ) recently ruled out the more granular option of nodal pricing but has kept zonal pricing on the table for its second REMA consultation.  

Dena Barasi, head of regulatory economics at Scottish Power, said locational power pricing would provide both long-run siting signals and short-term operational signals to network users.  

They decided to explore whether TNUoS charges could be used to provide the same siting signals but without exposing users to the risks of half-hourly pricing and “network delays that they can’t control and massively impacts the prices”. 

Under their OpTIC model, wider zonal transmission charges would be based on the difference in revenues (or costs) that generators (or consumers) could expect to see in an “unconstrained” national power market, representing the status quo, and a “constrained” zonal power market.  

These two scenarios would be modelled on an hourly basis over a five-year period and the average difference in revenues and costs in each zone would be translated into a £/kW capacity charge. In areas where the modelled zonal power price is higher than the national price, generators’ charges would be negative, meaning they would receive credits.  

By averaging out the differential over a five-year period, Barasi said their charging model would replicate the long-run siting signals provided by zonal pricing but smoothed out to become more stable and predictable. In doing so, the model would “cut off chunks” of the touted benefits of zonal pricing, which she warned would also bring “massive upheaval and very significant risks.”  

Short-run operational signals could instead be provided through more “evolutionary” changes to other market arrangements such as reforms to the Balancing Mechanism. Barasi said many of these changes would be needed anyway to implement locational power pricing. 

Network assumptions

Importantly, the modelling of national and zonal markets would assume there is “optimal” investment in transmission network – i.e. the expansion of the network to the point where the marginal cost of investment is equal to the marginal benefit. 

Barasi said this optimal network would primarily be based on the Electricity System Operator’s (ESO) Centralised Strategic Network Plan (CSNP).  

The ESO’s Holistic Network Design (HND) and its recently released HND Follow-Up Exercise represent the transitional iterations of the CSNP, the first full version of which is due be published in 2024/25. Barasi said the OpTIC model would effectively be a “bolt-on” to the CSNP process.  

The proposal document notes that the investments identified as beneficial in the CSNP would be assumed to take place when it is most efficient to do so: “In other words, if it was identified that it would be efficient for a particular network reinforcement project to be in place for the coming year then it should be included in OpTIC even if it cannot be delivered in practice for several years.” 

It therefore acknowledges that the optimal network could not be based solely off the current primary output from the CSNP, which “reflects the realistic date of delivery for new investments identified as being beneficial to the system”, adding: “A new or adapted output from the CSNP process would be required.” 

The document says the OpTIC model would solve a number of defects with the ICRP methodology, which “overly simplifies” the reality of “a complex system” and produces prices signals that are “hard to assess” for accuracy.  

The ICRP takes “no account of spare capacity on the network” and fails to value beneficial behaviour by consumers such as batteries, which can help to relieve constraints, it adds. The methodology is also likely to require continuous “cumbersome” and “time-consuming” updates to maintain its accuracy as the electricity system evolves.  

The proposal document says the OpTIC model does not require the same continuous updates or simplifying assumptions and takes full account of spare network capacity and the implications of demand behaviour.  

“Because it’s a modern economic optimisation model, it takes into account all technologies – demand and generation,” explained Dunn.  

As the OpTIC model would just replace the ICRP methodology, Dunn, who is a member of the CUSC panel, said: “It’s beautiful in its own way because it doesn’t change too much of what happens at the moment.” 

Barasi said it could therefore be introduced relatively easily and quickly, remarking that “it’s certainly easier and quicker than LMP is for sure.” 

DESNZ has suggested zonal pricing would take at least five years to implement.