Learnings

Outcomes

The MADE project has shown that there is significant additional value extracted through the coordination of multiple LCTs within a single premise. Both at a system wide level, and at a single property level there are tangible benefits, including de-risking the distribution network from unpredictable demand when assets are coordinated rather than operating individually.

Predictive controls are a key enabling technology for all of the above benefits of tariff optimisation and asset coordination.  Under the MADE project PassivSystems has trialled a sophisticated control system uniquely able to make the right quantitative trade-offs to underpin the complex decisions in controlling multiple low carbon assets simultaneously.

The modelling has demonstrated that current wholesale cost profiles and network charges, savings from peak shifting is a smaller component of the overall value stack compared to ancillary services revenues. The property demand and consumption patterns, as well as surplus solar available at the property, have a high degree of sensitivity on cost savings that can be achieved. 

The additional energy cost for providing ancillary services has a material effect of reducing the savings in energy costs from peak shifting. In some cases, this can be higher than the annual savings in energy costs, however this is more than offset by the additional revenue generated

The value opportunity from peak shifting and smart charging is low for customer types with low demand and low EV utilisation levels, and the value stack is heavily reliant on DSO services. For such customer types, if DSO service opportunities are not available, then there is little benefit from co-ordinated domestic flexibility at the household level. Moreover, if the EV is available for most of the time during the evening peak period, then with the EV by itself performing peak-shifting, a domestic battery would not be needed for such Low Demand consumer types (unless DSO services are available and pursued). 

The MADE concept offers material peak load shifting potential for the distribution network of between 35 and 40% reduction in peak loads on the network compared to optimised low carbon technologies optimised but in silo operation (based on half- hourly data). 

Imperial College has assessed the opportunities to deliver whole-system cost savings by utilising distributed flexibility based on the MADE concept are significant and increase with the level of uptake of the MADE flexible solution. In the 2035 horizon with an ambitious carbon target and high uptake of EVs and HHPs the gross benefits could reach £3.1bn per year, through allowing the electricity system to achieve the carbon target more cost-effectively, while at the same time reducing the need for high volumes of peaking generation capacity and distribution network reinforcements. The highest achievable net benefits, after deducting the cost of enabling residential flexibility through MADE, are lower (£2.1bn per year).

The net benefit is still considerable despite moderate levels of flexibility already being present in the system in the form of demand side response, large-scale battery storage and interconnectors. There is also a significant potential for distributed flexibility to deliver distribution network cost savings across different voltage levels and asset types, which can reach £200m to £500m of avoided annualised reinforcement cost.

Predictive controls that can optimise and coordinate asset behaviour play a key role in delivering best value from the assets to the consumer as well as negotiating patterns of behaviour desired by the local and national electricity grid.  The greater the level of coordination between the low carbon assets, the greater the savings in consumer electricity costs.

Time-varying tariffs can offer significant running cost benefits to consumers with MADE assets, particularly where the battery and heat pump can be coordinated to store energy in the right balance between the battery and the thermal fabric of the building and making the right decisions about waiting for available PV generation.

Even slight variations in tariff can reduce demand peaks, e.g. due to batteries delivering arbitrage.  These peaks can easily be mitigated by a smart control system, at only a small incremental cost to the householder, as long as the provision of cheap electricity is not significantly reduced.

With advanced controls it is expected that this flexibility and the associated benefits can be obtained without affecting customer comfort. This is essential if wide scale acceptance of advanced control of LCTs is to be achieved.

Lessons Learnt

Learning has been gathered throughout the project and is outlined below.

It should be noted that the trial was inherently at a small scale (five homes) and was only planned as an initial investigation into the concept. More robust statistical analysis would require a significantly bigger data set.   

Capabilities of coordinated control

• Predictive LCT controls that can optimise and coordinate asset behaviour play a key role in delivering value from the assets to the consumer as well as negotiating patterns of behaviour desired by the local and national electricity grid.  The greater the level of coordination between the low carbon technologies, the greater the savings in consumer electricity costs.

• Time-varying tariffs can offer significant running cost benefits to consumers with MADE assets, particularly where the battery and heat pump can be coordinated to store energy in the right balance between the battery and the thermal fabric of the building and making the right decisions about waiting for available PV generation.

• Even slight variations in tariff can introduce demand peaks, for example due to batteries delivering arbitrage. These peaks can easily be mitigated by a smart control system, at only a small incremental cost to the householder, as long as the provision of cheap electricity is not significantly reduced.

• There are a number technical challenges associated with coordinating control of assets. These range from difficulties integrating with proprietary systems to subtleties like not triggering “sleep mode” on certain vehicles.

• Traditional control of heating comfort has focussed on hitting minimum temperature requirements. However with the advent of negative electricity pricing, maximum temperatures must all be considered to prevent the homes from overheating.

• Under certain price conditions, the batteries were doing two cycles a day:

- Charge using very cheap overnight electricity, discharge to meet morning heating demand

- Charge prior to Agile peak and discharge over peak

      This is an interesting learning given that batteries are typically designed with one cycle per day in mind.

• Smart controls can effectively deliver both Secure and Dynamic Flexible Power services using the MADE assets, by pre-charging both the battery and the home in advance of the availability window. 

Benefits from coordinated LCT control 

• It is important to be clear on baseline behaviour when looking to establish the value of the services. The value of coordinated control should not encompass the wider value of individual smart control, but focus on the net additional value of coordination.

• Domestic flexibility does have the potential to provide a notable value opportunity. The Phase 1 desktop modelling work by Everoze Consultants showed possible savings of up to £260 per annum, per household. The technical trial completed in Phase 2 and Everoze attempted to validate their savings estimate but due to the small scale of the trial we should include some caution in the overall scale of per household savings. Equally not all premises are the same and this should also be factored in when measuring the value of this saving.

• In collaboration with PassivSystems, Everoze has identified that distribution networks can utilise the MADE concept by limiting loads to 33% of the 14 kW fuse limit at a property level without compromising household consumption behaviour and the savings that can be achieved (based on half-hourly average loads).