Exploring Time-Of-Use Tariffs

People that know me well, are aware that I’m unusually interested in electricity. Much of my career has been spent working in the field and I love talking about it. But when it comes to my own electricity use, I’ve historically been no more engaged than other people. I switch off my lights when I’m not using them, I pay my bills on time (by direct debit), and I would check my meter once a year, and about 4 years ago I switched to a green electricity provider. However, I’ve never given much thought to how much electricity I used or how much it was costing me.

Then, about a year ago, after a lot of phone calls, I had a smart meter installed, and was full of hope that this would finally change things. Unfortunately, it was a real disappointment. The display unit barely gives any useful information, and I can’t even use it in my flat (my meter is at the other end of my building). I did like the fact that I no longer needed to take manual meter-readings, but other than that I pretty much ignored it.

Then in August, I was at a seminar at Imperial College, and one of the students mentioned that Octopus Energy were trialling dynamic time-of-use tariffs, based on customers’ smart meter data. I figured that this would be a good incentive to engage with my consumption, and would hopefully save me some money (as a PhD student, I’m a bit more cost-conscious than when I worked full-time). It was a pretty simple process to switch, and by mid-November I was with Octopus. My smart meter even continued working, which I hadn’t anticipated.

What are dynamic time-of-use tariffs?

Most electricity customers pay the same rate per kWh, irrespective of what time of day they use electricity. Some customers get a cheaper rate for offpeak use, for example between midnight and 7am, and a higher rate the rest of the time. In the UK this is often referred to as Economy7, but the general term for such contracts is ‘static time of use’: static because the hours and the rates are specified in advance. I had previously considered switching to an Economy7 tariff (I have an electric boiler that heats overnight), but they were hard to get without a smart meter, and I had no idea how much I’d save, so I had never bothered.

A time-of-use tariff becomes dynamic when either the prices, or the times they apply to, change on a regular basis. In some cases it may be as simple as being advised certain periods for which a premium applies, encouraging users to reduce their use at those times. In the case of Octopus’s Agile tariff, every day at 4pm individual prices are set for each half hour the following day. Prices are capped at £0.35/kWh, but are generally much lower, occasionally even negative (meaning you get paid to use electricity)!

Saving Money

The good news is that I’ve significantly lowered my bills since moving to Octopus Agile, if I had to guess, I’d say by about 30–40%. My electricity bills weren’t huge before (my building has communal gas heating), but the reduction is still noticeable — £10–15 a month. That saving can be broken down into three parts:

• Half of that saving is through reducing my overall consumption. By monitoring my use, I realised that most of it was going on my water heater, which had been set to heat for 2 hours a day (6kWh), of which around 2kWh was being lost each day through poor insulation. For just £6 of insulation I was able to significantly reduce heat loss, and now heat for an average of 45 minutes per day (2.2kWh). I’ve also switched my bathroom and kitchen lights from halogen (120W) to LED (15W), and my living room lights from incandescent (80W) to LED (10W). Obviously there’s some upfront cost to this change of lighting — but it does feel good to see the reduced consumption.
• A third of the saving has been because I tend to use electricity away from peak times. In the UK, if you aren’t on a time-of-use tariff, you are assumed to use the same profile as every other residential customer, in other words, that you use most of your electricity between 4 and 9pm when wholesale prices are highest and the network is most congested. Because I use most of my electricity between midnight and 7am, switching to a dynamic time-of-use tariff means I now just pay the overnight rate.
• The remainder of the saving has been because of my flexibility. Rather than heating my hot water at 4am, I now heat it whichever half hours are cheapest over night (quite often there’s 3–5p/kWh difference between the cheapest and most expensive half hours between midnight and 7am). Also, if one morning is particularly cheap (and especially if prices are negative!), I’m able to heat more that day, and reduce how much I need to heat the following days. At the moment I am still having to adjust the boiler times manually, but at some point I probably will move to an automated/remote switch. Finally, I do check prices before scheduling my washing machine.

I should make clear that a dynamic time of use tariff like Octopus Agile won’t be for everyone. If you use a lot of electricity between 4 and 7pm, you’ll be much worse off on a time-of-use tariff. If you need the price certainty, stick to a fixed tariff. And if you find having to track half-hourly electricity prices too much effort, maybe wait a year or two until you have devices that are able to do the scheduling automatically. But if you’re like me, it is a great way to save money, reduce national electricity demand when the grid is most congested, and test out flexibility-promoting mechanisms that will become more and more important as intermittent sources of electricity (especially renewables) become the majority of our power supply.

The challenge of accessing useful data

None of what follows should be taken as a criticism of Octopus, who I have always found incredibly helpful, and the level of data they provide is much better than any of their competitors. But I do feel there is significant room for improvement, if dynamic time of use tariffs are to become more widespread.

• Octopus provide an web API, which allows customers to access their half-hourly consumption (up to the previous night), and their half-hourly prices (up to the following day if called after 4pm). This is great, and should be mandatory for all suppliers. Most other suppliers currently make it impossible to see your consumption data until you receive your monthly bill!
• On the back of Octopus’s API, I was able to build a website to analyse my consumption and price data. And even better, one user built an iPhone app, Octopus Watchdog, which allows users to view their consumption and price data in a user-friendly interface. This is great!
• However, using these APIs is not easy, and I find it remarkable to suppliers like Octopus don’t make it easier to view consumption data on the website. (I’m sure this isn’t far off in the case of Octopus, and will be a huge improvement — if only other suppliers, and indeed the regulator, would make this the norm.)
• I find it very difficult to compare my bills now, with what it would have been on a fixed tariff. I mentioned above that I thought I was saving 30–40%, but I’ve got no way to be sure. Fixed tariffs tend to be for a whole year, including a winter with higher prices and a summer with lower prices. In contrast, when I look at my prices for tomorrow, I am seeing them for one day in February. I would love the ability to see what their equivalent fixed price tariff was for tomorrow, for the rest of the month, and for subsequent. This would not just give me a sense of how much I was saving, but also allow me to forecast how much my prices were likely to change over the year.
• It is definitely something that is easier said than done, but I would love to have an estimate of what the half-hourly prices were likely to be a few days out. I base the amount of time I heat my hot water each morning on the cheapest prices that morning, and my expectation of what the cheapest prices will be the following mornings. If I could make this more accurate, I could save money.
• Smart meters capture your electricity usage in real time, however, it is only sent to your supplier daily. This means that consumers have three choices for accessing their data — getting it the following day from the supplier (as I currently do), getting it from the smart meter (as I can, but very limited data), or installing a home energy monitoring system. Unfortunately, at the moment, home energy monitoring systems are expensive, and I believe unable to be used if, like me, your smart meter is too far away from your flat. I believe this is something that the government will need to address if they want more customers to be more responsive.

A few other thoughts

• At the moment the prices that Octopus charge Agile customers are based on market prices at around 4pm the day before. This seems a reasonable decision, giving customers some level of certainty, and minimising the gap between the price that customers are charged and price that Octopus have to pay for that electricity.
• However, this approach prevents any opportunity for customers to respond to changes in prices between 4pm the day before, and the half-hour itself. This will be particularly useful as devices like electric vehicle chargers and hot water systems become automated, and can easily adjust to changes in supply and demand.
• On the other hand, I believe some customers would like the ability to hedge portions of their demand. For example, a customer that knew that they were likely to use power most days between 6pm and 8pm, might choose to lock in a price for some of their demand. Similarly, I would feel more confident not heating my water one morning if I could lock in a price for the following morning.
• I believe that as renewables become a large part of our electricity mixture, flexibility of demand is going to become more important, and more valuable. However, as more people take up that flexibility, it will get more challenging to manage. For example, if Octopus and other suppliers advise customers that the price will be particularly low at 3am, enough may move demand to that time so that it ends up being high, costing significantly more. I don’t have the answer, but I am extremely interested in this challenge.
• Finally, technology and markets are changing rapidly, and I believe Octopus is absolutely doing the right thing in being transparent, building APIs that allow for the system to be used with other platforms (eg ITTT, Hive, Nest), and involving customers in the discussions.