Optimisation is a balancing act, in which economic considerations are as important as technical ones. Image: Chargesync.
So let’s imagine for a moment that we’ve constructed a network of embedded batteries across the country which can all be controlled independently. What would you tell these batteries to do, and how could you be sure that what you’re telling them to do is correct? Which market signals do you listen to? What is your primary driver for actions?
Voltage? Frequency? There can be times when this might be useful, but can you justify the value of your battery on the basis of following frequency and possibly being paid by the grid?
Since placing a battery onto the grid is an economic investment which requires capital to purchase the box we need a strategy which will extract the maximum value from the battery in its position on the grid. The problem is complicated because there are many potential sources of value. Do you simply sell the control of your battery to National Grid, or the DNO? Do you try to control the battery in real time? Do you respond to market pricing in the form of time of use tariffs? Or do you try and use your battery in the balancing market? Are the services you offer mutually exclusive? Or can we partake in more than one service at the same time? Trying to solve the problem all at once is probably not possible, so the approach we take is to break the value chain into pieces.
Economic optimisation on a forward-looking basis
Here at ChargeSync our approach begins with an economic optimisation of the batteries on a forward looking basis against the next 24 hours of market prices and time-of-use tariffs. Once this is done we will devise an optimal instruction set for every battery under our control. We believe that maximising the economic potential for the batteries given market prices is a critical initial step as this provides us with a price against which to sell further actions, or change the strategy. Without performing an economic optimisation of the device in your own control you are unable to tell what the value/opportunity cost of providing any other services would be.
As well as an optimal strategy our economic optimisation provides us with a method to value changes to our strategy, and hence economically price and sell services into the spot market. Once the optimal strategy is computed changes to that strategy (e.g. withdraw when you were going to inject) can be evaluated, simply, if the value of the service (what you’ll get paid for the service) will exceed the value lost from the initial despatch, then we would perform the service, if not then we won’t. Without this methodology you either accept the value offered to you for grid services, or don’t, but no real criterion exists to make that decision. The key thing here is that current grid services alone are unable to underwrite the value of a battery, and nor would we expect them to.
The key thing here is that current grid services alone are unable to underwrite the value of a battery, and nor would we expect them to."
So if you want to maximise the value of your battery based assets, and know how to price your services then we believe a market based optimisation is the logical stepping stone. There are complications of course. For instance, if you are participating in grid services then when do you need to commit to the service, and what is the cost of failing to provide the service you promised? These are all trade-offs that need to be evaluated by your optimisation algorithm and result in a changed, or unchanged despatch of the asset.
It’s worth mentioning that this process needs to be repeated at regular intervals. The optimisation needs to respond to real time events in the market, and to the evolution of charge levels within the battery. We expect price, weather and demand to be exogenous variables in our calculations.
Finally we should mention Distribution Network Operator (DNO) grid services from batteries. Our aim is to provide a real time price (a shadow price in economic language) for the marginal unit or charge or discharge. This shadow price will allow us to create a price for any opportunity. This means if a DNO wants to withdraw discharge batteries for a particular street in Manchester then there’ll be a price for that. Local opportunity pricing can and will (in my view) create a whole new market for energy on the low voltage grid.
It’ll be interesting to see whether others look to follow our approach. Most folks seem to be plumping for grid services as a revenue stream, but our view is that just isn’t… optimal.