First they came for your kettles… household control and demand response

Demand response is becoming a hot topic in energy circles, how we can help to level out the peaks and troughs in the energy system by utilising flexible demand. Energy demand changes hugely over the day, from being very low in the middle of the night to peaking on weekdays usually around 6pm. It also changes in less predictable ways, responding to co-ordinated mass activity such as the ad breaks in popular tv programs. As Wimbledon has just finished, the graph below shows how electricity use changed during Andy Murray’s victory in the finals in 2013.


The typical response of the electricity network operator has been to increase supply to cope with these types of events, making sure that there is enough flexible supply able to deal with the peaks. This can be technologies such as pumped hydroelectric power, which can respond as fast as 16 seconds to requests for more generation. However, there’s two big reasons why just increasing supply isn’t the best option.

  1. Cost – Having generation which you only use at peak times is incredibly uneconomical, and requires expensive and complex markets to ensure that the capacity is available.
  2. Environment – In the recent capacity auction many of the winning bids were from incredibly dirty diesel generation. There are also frequent calls for natural gas, including from fracking, to be used to cover peak demand.

One better solution, many people would argue, is to reduce the impact of co-ordinated activities by getting households to shift their demand. This can help to level out the peaks and reduce both the cost and carbon emissions of the electricity network. In a system where excess electricity is generated during the day from renewables, ideally demand could be shifted into these periods. The question is how to make this happen. I’ll try to outline some of the main challenges to making this work, and make some amount of judgement on whether this seems to be a realistic possibility.

Exposing consumers directly to (more) variable electricity costs

Currently electricity bills give you one (or maybe two if you have an economy 7 tariff) electricity price no matter when you consume your electricity. Some would argue that consumers should be exposed more to the real time prices of their electricity, which change quite significantly during the day. The argument goes that if consumers were given prices by the half hour, or every five minutes, they would be willing to shift their demand to achieve savings in their electricity bills. This seems to be complete fantasy. I’m all for consumer engagement with their electricity usage, it’s kindof my thing, but this seems several steps too far. Sure, a minority of nerdy consumers would love this, but the vast majority of people have far better things to do with their time. Energy regulators have enough of a problem getting people to engage once a year with their electricity supply, getting them do it every time they want to make a cup of tea would be near to impossible.

There may be a possible world where electricity prices go up between 6 and 7 in the evenings, and this is commonly known among consumers. This might help to combat some of the issues at the very peak, but it wouldn’t be able to cope with more complex forms of shifting, such as based on when renewables are generating excess capacity. It also goes against the current perceived wisdom that electricity tariffs are too complex, and adding another layer of complexity would make comparing between offerings more difficult still.

‘Smart’ technologies and controlled homes

Some companies (e.g. Tempus energy) argue that the solution is for a middle-man (apologies for the gendered term) to exercise a level of control over peoples’ homes and manage the electricity usage on the their behalf. This is quite a fashionable idea, and one which is attracting a significant level of financial investment. However, the proponents of this type of system seem to be unaware of and unwilling to deal with important issues of control. In order for the middle-man to make a profit after investing in the necessary technologies to allow this type of system to operate, they need to be able to dictate when energy is used. The consumer on the other hand will only be willing to sacrifice so much for marginally cheaper electricity prices. If a household wants to wash and dry clothes in the middle of the day because they just need to, then there’s no chance that they will accept a third party telling them they can’t. Sara Bell, CEO of Tempus Energy, came to speak at Oxford recently and seemed entirely unwilling to accept that this might be an issue. This does seem slightly strange given that this is the most contentious area of household level demand response.

The middle-man concept isn’t entirely pointless though. There are some technologies which could be managed. We can call these the ‘background’ processes of a house, the ones which are pretty constant and most people don’t ever think about. This includes space heating, cooling, fridges, freezers and water heating. (For sake of comparison, ‘foreground’ processes could include kettles, ovens, washing machines, dishwashers, or televisions). These ‘background’ processes are where household level demand response could really work, such as heating your water when electricity prices are low or switching off the fridge temporarily (fridges don’t run constantly, they can switch off for half an hour without anyone noticing). As long as consumers are always assured they will have the services they need as reliably as ever, all of the problems of control will disappear. The important question then becomes whether simply controlling these few services provides enough of a business case for actually installing the necessary smart home technology.

The future: artificial intelligence

Channeling my inner Elon Musk, here beings an indulgent list of vague questions about the future of technology.

Could an AI control our energy usage? Could AI solve the problems of control by being aware of our preferences and able do the cost benefit analysis of demand shifting for us? Could an AI enhanced central network operator be able to optimise the system so that this type of consumer demand shifting is no longer necessary?

In conclusion

Lots of the discussion about demand response at a household level ignores the issue of control, and often just assumes that once everything is ‘smart’ these problems will resolve themselves automatically. This won’t be the case. People will still want to come home after a day of work, make a cup of tea and cook dinner regardless of how smart their appliances are. Unless companies involved in this market think about control, and only aim at shifting ‘background’ processes, there is no way that this type of business model will ever work.

First they came for your kettles… household control and demand response

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s