Learnings

Outcomes

A set of learning objectives were defined at the start of the CLDS project. The learnings summarised below will help us gain more understanding of the possible future of flexible energy systems and the implications of our transition into the future role of the DSO. 

1.       Whole-system value of distributed flexibility

Flexibility from DER saves £15.1 bn pa (net) in the deep electrification scenario. This can be achieved by investing £6 bn pa in flexibility assets.

The value of CLDS flexibility is sensitive to heat decarbonisation strategies. In Hydrogen & Hybrid Heating pathways, the value of flexibility is 7.2 to 7.7 £bn pa compared to £15.1bn in the deep electrification scenario.

2.       Breakdown of the value of distributed flexibility

Around 90% of the whole-systems savings from deploying distributed flexibility are related to the electricity energy supply chain and between 80-90% of the total savings are on energy system CAPEX. 

However, only around 30% of the total savings are in the reduction of required distribution network capacity. We have seen a significant reduction in costs related to generation assets and other systems such as hydrogen, heating and carbon removal. This shows the value of the CLDS is beyond the interests of the distribution system. 

When optimising the deployment of distributed flexibility resources, a whole-systems approach provides more benefit as compared to a DSO-only or non-DSO approach. This will require the DNO to build more infrastructure to enable distributed flexibility to support the broader energy systems.

3.       Flexibility and distribution capacity

Distributed flexibility can reduce c.17%-33% distribution capacity requirement. The capacity reduction varies depending on the scenarios, regions and network types (urban and rural systems). The highest reduction is observed in the deep electrification scenario.

Without flexibility, deep electrification of heat and transport leads to 260 GW electricity peak demand, which is much higher (three times more) than the present capacity. Although distributed flexibility can deliver great savings to the distribution system, significant network reinforcements are still expected. 

To meet increased peak demand and use flexibility to support transmission system operation, around 66%-83% of the capacity needed should be provided through network reinforcement.

4.       Customer investment and system investment

There will be a trade-off between the costs directly seen by customers and the wider energy system costs which will be impacted by customers’ decisions. For example, rather than buy a heat pump sized to meet peak heating demand, customers may opt for a lower up-front cost option of a smaller heat pump in combination with resistive heating to meet the peak heating demand. From the whole-system perspective, this could be the least-cost (with a whole-systems saving of around £1-2 bn pa) solution compared with only using a larger heat pump. Although the former will require more electricity system capacity (including distribution network) to accommodate customers’ choices. The savings in heating appliance costs exceeds the increased system costs, including distribution network reinforcement cost.

5.       Value of local energy markets

Our initial analysis of modelling of the LEM and network service markets shows that DERs’ benefit from the LEM (price-driven) could be 22 to 63 times that from the network service markets (contracted).

Our multi-agent multi-market modelling further shows that, amongst all markets and services considered, LEM appears to be the most beneficial to the DSO, provided appropriate price signals are instigated. This can lead to a substantial decrease in DSO costs for flexibility, and in some cases avoids the need for a DSO service altogether – up to 100% cost savings on DSO flexibility services. In particular cases, we observed that introducing ToU tariffs and adjusting DUoS charges reduced peak demand and overall demand for flexibility; this reduced DSO flexibility requirements by 36% resulting in a 55% reduction in DSO flexibility costs.

6.       Local characteristics in local energy markets

The modelling assumes the customers are only motivated financially to participate in the LEM. However, the benefit will reduce if the customer’s objective shifts. In a case, we observed a 50% reduction in DSO’s benefit if CO2 reduction has been considered in customers’ decision-making when trading in the LEM.

DSO’s benefit from the LEM is also highly dependent on the availability of renewable generation. In some cases, we observed an increase in DSO’s contracted flexibility cost up to 68% when renewable output is low. During these periods, price-driven flexibility contributed by LEM activities is low and more energy needs to be imported from wider transmission networks.

DSO’s benefit from the LEM is also related to the level of customer participation. Although a higher level of customer participation in the LEM will lead to a lower requirement of DSO flexibility capacity, the DSO’s cost reduction in flexibility is non-linear to the capacity reduction. This is because the LEM activities will influence the prices of DSO flexibility and in other markets.

7.       Interactions between markets

Our multi-agent multi-market modelling showed that suppliers’ market for contracted flexibility will help reduce DNO’s need for flexibility. This can be directly influenced by DUoS charges. When DUoS charges rise, suppliers will offer competitive prices for flexibility services during peak hours (where the wholesale price is high) to reduce their loss of making sales. Although suppliers’ needs compete with DSO’s needs for flexibility, the net benefit is still positive as their objectives are aligned during peak hours and therefore their actions are synergetic at those times.

Our modelling indicated that ESO’s market doesn’t necessarily have a negative impact on the DSO. Theoretically, ESO’s flexibility actions for over-frequency response support can be detrimental to the DSO as additional flexibility needs to be secured by the DSO as a preventive measure in case the ESO calls for distributed flexibility in a congested network. However, the ESO is also naturally incentivised to procure flexibility services from a non-congested network (frequency response doesn’t need to be localised) to avoid competition.

8.       Impact of flexibility on losses

Distributed flexibility’s major benefit to the DSO is investment deferral due to peak demand reduction. We also found, from the operational perspective, utilising flexibility can also reduce network losses. Depending on the type of DER, distributed flexibility may also increase or reduce the peak utilisation of the EHV network.

9.       Market design and clearing mechanism

The multi-agent multi-market modelling demonstrated that competition between the markets will drive up the clearing prices, whereas the competition within a market will drive down the clearing prices. 

Our modelling also validated the effects of different clearing mechanisms on the clearing prices. Pay-as-bid (PAB) produces the highest clearing prices and therefore the highest revenue for the flexibility providers. This can be used to attract providers to emerging markets but is also prone to market manipulation. Pay-as-clear (PAC) and Dutch reverse auction (DRA) can be used to produce more cost reflective clearing prices as compared to PAB. PAC is easy to understand and widely used.

Lessons Learnt

The value and role of flexibility in the 2050 energy system

Harnessing flexibility from distributed resources saves the cost of future GB energy system by £15bn/year (dependent upon the specifics of the system). 30% of the savings are in electricity distribution network costs, 70% are in electricity generation Capex and Opex, heating, hydrogen, carbon infrastructure. Significant flexibility is deployed on the Northern Powergrid networks and it can reduce distribution capacity requirement by around 30%-40%. Flexibility is used more frequently for balancing but during times of peak demand it is also used for peak reduction i.e. it the deployment of flexibility follows a combination of renewable generation and network demand.

Whole system or local optimisation

An approach that seeks to use distributed flexibility to minimise the cost of transmission and distribution networks acts to increase the annual cost of the whole energy system by £2bn/year compared to taking a the whole-system approach for using distributed flexibility. 

Optimising for the whole system reduces overall costs but increases peak demand on the Northern Powergrid network by about 20%, compared to optimising transmission and distribution costs. i.e. the distribution system can be sized and used to minimise the whole systems costs which may mean that more network is required.

Distribution peak load sensitivity to sources of flexibility

We have modelled how the network peak load depends on the source of flexibility and considered the impact should any of the following not thrive: smart EV charging; smart thermal storage; and batteries. The conclusion was that if one of these fails to take off then the gap will be taken up by the other successful technologies and the benefits from flexibility will still be largely achieved with a significant reduction in peak loading. However the impact may depend on location with some technologies (eg smart EV charging) impacting all voltages (LV up) and some (eg large scale battery storage) largely impacting HV upwards. One flexibility technology can be substituted by other technologies with most of the benefits captured e.g. batteries and thermal storage can be alternatives for smart EV (and vice versa).

Sensitivity to offshore wind capacity

If the country delivers 80GW of its 120GW target for offshore wind, the shortfall will be filled by CCGT with CCS, onshore wind and solar. This reduction in offshore wind will not change the sources of flexibility on the distribution network but it does reduce the value of flexibility by 6% although the value is still significant. 

Electrical heating options and system costs and benefits

For areas relying on electrical heating, heat pumps will need to be either ‘super-sized’ to deal with 1 in 20 year multi-day very cold spells, or resistive heating will be needed to supplement a “normal-sized” heat pump. The customer choices are to either invest in a larger HP system (higher direct cost) or to use supplementary resistive heating with higher system costs due to the need for more generation capacity and more network capacity ie there is a trade-off between the cost of heating appliances and the cost of the wider energy system. Overall it is cheaper to use resistive heating to deal with infrequent extreme weather rather than to ‘upsize’ the larger heat pump for such events. However, the impact is that distribution peak load increases by up to 30% if resistive heating is used alongside heat pumps, and more distributed flexibility is needed to reduce the electricity peak demand

Low Carbon Technology (LCT) growth investment planning tool

We have enhanced the existing business-as-usual LCT-growth planning tool so that it is regionalised for the Northern Powergrid network and so that it can model the impact of flexibility. We have used this more sophisticated tool to generate and test our HV/LV load-related costs for our ED2 business plan.

Flexibility markets

Early findings from multi-agent modelling of participants in flexibility markets indicates that the market clearing mechanism (eg Pay-as-bid, Pay-as-cleared, Dutch reverse auction) affects bidding strategies of the participants and consequently the flexibility capacity bid into the market and the price paid for the service.