In the last in this series on developing your Demand Response strategy, I want to talk briefly about some of the analytical models that will underpin your success.
Throughout this series, I have highlighted, both directly and indirectly, the need for a strong analytics capability. This required capability is not just in terms of the technologies, but of the know-how required to design and continuously improve them. My views on the design of these models continue to change since I was first exposed to them four years ago.
There are five I want to bring your attention to:
- Knowing how to place a value on the demand you have cut
- Knowing when, where, and how much demand cut you need
- Knowing how to calculate the financial incentives to pay customers
- Knowing how to calculate the actual demand that was cut
- Knowing which customers to target
There are other insights required that need strong analytic support, but these are the five that I believe contribute the most to success, or failure, of a demand response strategy.
Knowing how to place a value on the demand you have cut
During your initial analysis, you need to place a dollar value on a NegaWatt (NW). Any utility investing in generation, transmission, or distribution assets would have models that help it calculate a cost per MW; therefore, you can derive the value you place on a NegaWatt (NW). You should separate Generation assets from Transmission and Distribution as the latter may only deliver value in certain locations on your network. In places that have wholesale markets, again there will be a per MW or MWh value that can be calculated by bidding into the wholesale market.
Knowing when, where, and how much demand cut you need
Now you have placed a dollar value on your NW or NWh; you need to determine your target NW, or NWh. Again the approach will vary by market, but when you have a target you need to understand how to achieve it. You need to determine when is the best time to call an event. The optimal duration of the event. The amount of time you have to call an event. You also need to know whether this applies to an entire network or a particular location of the network. Finally, you need to know how many times a year these events need to be run.
Knowing how to calculate the financial incentives to pay customers
I touched on this in a previous post. There are various models such as Peak Time Rebates, Critical Peak Pricing, Real-Time Pricing, and some emerging thoughts around fixed rate subscriptions. You cannot simply apply the number of events with the required NW to calculate how much this will cost you. For example, calculating peak time rebates, is not an exact science. Models such as 3 in 10 to create a baseline, or its many variations, tries to predict what the customer would have consumed without the incentive being in place. While this is a commonly accepted approach, it does mean you will pay some customers an incentive that did not do anything, known as free-riders. You also need to factor in that a percentage of customers will not take part in an event, yet you still incur costs in trying to engage those customers to get them to participate.
Knowing how to calculate the actual demand that was cut
You may ask, “if you have calculated the financial incentives to pay each customer, then surely you now the total demand cut?”. Not necessarily. I am aware of work being done to bring the two together into a single model, but often you will have a separate model that calculate the overall demand cut for the event. One strong recommendation. Engage an independent expert to help develop, or at least, review the design of this particular model. These are complex models that need a series of assumptions to be made when considering a range of variables. If your demand cut calculations will be subject to third party scrutiny, such as by a regulator, then it is worth the initial investment to get this checked out.
Knowing which customers to target
You now know how much it is going to cost you to engage a customer, and how much incentives you need to pay. So you need to minimize your engagement costs and to pay out unnecessary incentives. Using insights such as load profile data, whether the customer has signed up to digital channels, and how often they interact through those channels will all help you design an engagement plan that gives you the greatest chance of realising your demand cut targets at the lowest cost.
I hope you enjoyed this recent series. In the coming weeks, I will be publishing a new series giving my views on the Australian Energy Market Commissions (AEMC) reforms to Australia’s National Electricity Market (NEM), known as the Power of Choice. Intended to meet the needs of the energy consumer over the next 15 to 20 years. I hope it will be useful to those involved in the NEM, and informative to others around the world as an indication of developments that may happen in other markets.