Is Flexibility the new Efficiency?

sunshine On 7 June 2017 around lunch time the UK generated the majority of its electricity from non-fossil fuels. That was the first time, the first time ever. It also was the time when wholesale prices for electricity plummeted to a tenth of their normal level.

It's a sign of things to come. Renewables are falling in cost so rapidly that cost is no longer their problem - it's becoming our problem. Low costs that is. Once built their electricity at time of generation is practically free (i.e. one doesn't save a penny not to produce it - very different from plants with fuel costs). When the sun shone and the wind blew on 7 June, all that renewable electricity displaced more expensive (yes, more!) fossil fuel electricity. Wholesale prices head towards zero temporarily.

What does that mean for us, the 'just about managing' electricity users? Sadly, as things stand, nothing at all. We cannot get our hands on that super cheap electricity at the time, nor are we likely to see it in lower bills over the year. There are two reasons for that: in order to benefit from temporarily low prices one would need to know how much you were using at that time. With electricity meters being read randomly every six months that is impossible to tell. Hence, the need for smart meters, which could in theory give us a different price every 30 minutes.

Before you jump for the opportunity to get your hands on these low cost periods, there is inevitably a flip side. The reason we won't necessarily see lower bills despite lower cost renewables is that there are of course times when they are not available. And those times can become much more expensive. You could blame coal and gas power stations for the highest prices, but that would be missing the point. They are faced with a tough prospect: all the cheap renewables are eating away at their market. Fewer and fewer hours remain when coal and gas plants can operate. With fixed costs stubbornly, well, fixed, the price for the remaining electricity they produce goes up.

So how do we get the cost down? Easy: if we all continuously avoid the most expensive periods, some of those old power stations can enter their well deserved retirement, overall prices fall, as do emissions. Even better, at times with lots of sun and wind we can even use more electricity. It is no longer about efficiency (using less), but flexibility (using wisely).

Is that a realistic prospect? Would you reschedule when your appliances run? Are you even able to? Smart technologies might help, but much of our lives don't give us much room for movement. The University of Oxford tries to understand the time pressures in our lives and how they shape our energy use. Take part and you could help bring down the energy cost for all (and with a bit of luck win a year free electricity for yourself - that really brings down the cost).

Register here

Why is so little know about electricity use?

Authoritative looking pie charts like this one use_pie

give the impression that we know who is using how much, for what and when. In fact these figures are rough estimates of averages. They are often based on assumptions and a few measurements in small studies.

These studies are very limited in their ability to

  • capture the timing of electricity use – be that the time of day or season of the year,
  • give some sense of the diversity of users – some use a lot, some very little and it depends a lot on circumstances,
  • highlight the scope for changing electricity use – which uses are essential, which ones would go unnoticed if we shifted them a bit?

There are two main reasons why these details are poorly understood:

  1. We didn’t really need to know (until now)
  2. It’s not that easy to find out

1) We didn’t need to know

Fossil fuel power stations are very good at responding to our electricity demand. When you switch a light on, somewhere a power station will increase its output by a small amount. This keeps supply and demand in balance – second by second.

So long as the system operator can roughly estimate how much electricity we require collectively over the next few hours, this approach works well. No need to know what this electricity is used for.

However, as wind and solar generate more of our electricity, it becomes harder for supply to follow demand at every instance. One hope of system operators is that demand itself could become flexible. For that to happen it does become interesting to understand what ‘causes’ all that electricity use, so that we can find effective ways to avoid or shift it.

2) It’s difficult appliances

The second reason why we know little about electricity uses in households is that it isn’t that easy to find out. A 6 monthly meter reading certainly doesn’t tell us much.

Various approaches have been applied.

Appliance stock models: One can try to estimate how many appliances we have in our homes from retail sales figures. However, that still leaves us guessing when and how much we use them.

Household instrumentation: A more direct way to find out is to measure appliances directly in the homes. That means every light, every appliance from the vacuum cleaner to the cooker must be ‘wired up’ and measured. This process is very intrusive for the people in the household and it is also quite expensive. So expensive that no more than 50 households can be observed in this way – nowhere near enough for statistically robust research.

Demand disaggregation: A novel way to reduce instrumentation cost is to take a high resolution reading at the household’s main meter and try to recognise which appliances are in operation based on their characteristic profile. Fridges, for example, have a typical on-off pattern, washing machines can be identified by their spin cycle and even smaller good give away clues when observed closely.

While these approaches are getting better at finding out what the appliances are doing, they still don’t tell us much about what we are doing. The television might be on – but is anyone watching it? All the lights are on, but is anyone actually at home?

If it is flexibility we are after, then these differences matter. This study will take a very simple approach to find out: we ask people what they do. Here is how…

Measuring and Evaluating Time- and Energy-use Relationships (METER)

WhatAreYouDoing This 5 year EPSRC fellowship will use smart phone technology to collect activity information from UK households. In combination with electricity use profiles these data will give us a new perspective of what we use electricity for.

Read the full Case for Support

METER addresses a fundamental research question: “What is the temporal relationship between electricity consumption and household activities?”. To date this relationship is still poorly understood. METER will address this gap by collecting electricity consumption data in parallel with time-use information using adapted smart phone technology.

A detailed understanding of ‘what electricity is used for’, especially during peak demand periods, is important in addressing emerging system balancing challenges and to develop appropriate policy frameworks and business models leading to the cost effective integration of low-carbon generation.
At present electricity is supplied based on a ‘predict and provide’ paradigm – so long as we can forecast ‘how much’ electricity is required at any one time, the fleet of mostly fossil fuel based plants can be scheduled to deliver. Little knowledge about the end-uses of energy has been required for this approach. With low carbon sources, such as nuclear, solar and wind, more flexibility may be required from the demand side. Understanding the end use activities supported by electricity becomes more important when seeking to reduce or shift the timing of consumption.
Studies attempting to measure electricity use at the appliance level have so far been limited in their scale by the cost and complexity of instrumentation. The absence of statistically robust consumption data has been noted as limiting the UK’s world leading research in this area.
METER develops a new approach to collect electricity consumption in parallel with time-use information. Smart phone technology, developed by colleagues at Oxford, will be deployed to measure electricity consumption at 1 second resolution and ask participants about the activities they undertake at critical times of the day. The use of smart phones allows this process to be performed at unprecedentedly low costs, such that over 2000 households can be included in the study. This scale is important, because electricity uses are highly diverse and only a sufficiently large sample allows to develop statistically significant evidence for researchers and policy makers.
The concurrent collection of time-use and electricity consumption can improve the accuracy of time-use research and provide new insights into the use and timing of electricity consumption and its relationship with household activities. The data and the analytical tools developed by METER will provide much needed insights into the timing of electricity uses, which can underpin a wide range of future research priorities. Among them are emerging energy system balancing challenges and broader policy challenges relying on statistically robust information about the relationship between energy use, demographics, lifestyles and their transitions over time.
Findings and insights from METER trials will become publicly available as part of a public outreach campaign, including interactive online tools to explore how Britain uses its electricity and what the public can do to support the transition towards a lower carbon future.

Tesla Powerwall –
is it worth the money?

Tesla Powerwall received great media attention and many people have asked me since: is it worth spending almost £2300 on it?

Here is a short proof why you should not buy it if you like money, but you should if like like mankind.

Let’s make your ‘private’ economic case first by assuming the best possible case for storage:
1) You already have PV to match the Powerwall: 3kW (about 12 panels)
2) You use exactly as much electricity as your PV produces: 3,000 kWh/year (just below UK average)
3) But you never use electricity when the sun shines (unlikely, but we try to create the worst possible case – or the best from a storage perspective)

From this starting point, storage could be great. You currently:

  • have to export all your PV
  • 3,000 kWh for which you get ~4 p/kWh.
    You earn £120 per year.

  • import all your electricity back from the grid
  • 3,000 kWh at ~12 p/kWh.
    You pay £360 per year.

    Your cost: £360 – £120 = £240 per year.

Now with storage:

  • no more export. The power wall charges up when the sun shines. You no longer get those £120.
  • no more import. All your use is now met from the Powerwall. You no longer pay £360 for electricity.

Your saving: £240.

£240 is the theoretical maximum under the most favourable conditions. More just is not possible. As an investment, with this best possible assumption, over a ten year period, even with a very modest ‘social interest rate’ of 3.6% your Net Present Value is negative (around -£280). So, if you like money, do not buy it.

But is that all? Above we used electricity cost and export tariffs (note that the Feed-in tariff is irrelevant here, because – despite its name – you get that irrespective of whether you consume or export).

The cost of electricity is a poor measure of its value. Peak time electricity, for instance, is far more expensive to generate than night time electricity. Similarly, if your neighbour is using electricity, he would probably be very happy to have yours for somewhere between 4p and 12p. Yet, if he doesn’t need it, and nobody else wants it either, then it could soon pose a huge problem for the network operator. Enough of a problem that it could be worth another £50 per kW per year. The Powerwall is 3 kW, so that would be an extra £150 per year. Add to that the extra power stations we can avoid building – another £150 perhaps – and the case starts to turn.
For you to get that money, we have to make some fundamental changes to our billing system, or perhaps come up with novel business models to make money out of providing such ‘system services’. Until we have those, take comfort from the fact that with your Powerwall you will be helping the system and mankind – not your wallet.

Is a 16 fold increase in PV bad for Oxford?

April 2015. BBC Radio Oxfordbbc_radio_oxford

Distribution Network Operator SSEPD is concerned about the fast rise of PV power in Oxfordshire. BBC Radio Oxford wanted to know what we should do. Here I talk about network reinforcement, storage and demand response – well, Howard Bentham was most interested in storage – and whether I like the look of PV… all in 3 minutes.

Combining time-use and energy research

9 March 2015, WholeSEM workshop, University of Surrey

At the WholeSEM workshop I proposed how combining time-use and electricity data collection can fill an important gap in our understanding of ‘what we use electricity for’.

The audience response was very positive. This could develop into a transformative area of research!


Keep it simple: time-of-use tariffs in high-wind scenarios

MattIf wind and solar are to become major contributors to our electricity mix, we need to learn to make better use of them. One approach to ‘teach us’ would be to make prices dynamic: cheap when its windy, expensive in a wind lull. That could encourage ‘efficient’ behaviour, but might also get quite confusing. We already complain that there are too many tariffs. Do they have to change every hour?

This paper argues that a lot could be achieved with relatively simple and static ‘time of use’ tariffs, which merely discourage use during the usual peak periods. Somewhat surprisingly, we don’t all have to become dynamic traders. Reducing peak demand remains to be a valuable contribution. In fact, it becomes more valuable with more wind on the system.

IET Renewable Power Generation paper (open access)

The role of hydrogen and fuel cells in providing affordable, secure low-carbon heat

H2FC Supergen. 2013-2014. Imperial College, UCL

Hydrogen has gone in and out of fashion, especially in transport. But what about heat? Heat may not be as sexy and transport, but in the UK we need a lot of heat. Households use four times as much energy for space and water heating as for electrical appliances. And most of our heating relies on natural gas.

Renewables can help to decarbonise electricity and one popular proposition has therefore been to electrify heating as well using efficient heat pumps.

This, however, brings with it one serious problem: we need most heat at the very time when we already demand the most electricity, typically on a cold and dark winter weekday evening between 5 and 6pm. Adding another load to this ‘peak demand’ is very costly – it requires extra power stations and extra network capacity.

This study showed how complementing heat pumps with fuel cells, which generate heat, but conveniently also produce electricity at the same time, would be very advantageous in this scenario. A combination of fuel cells and heat pumps can balance out the increase in electricity demand and avoid costly reinforcements.


Publications: IJHE paper (open access), White paper

H2FC At the launch event in London

How to understand energy demand

Demand response – who is responding?

DSRTechnology Strategy Board, Lead Partner Kiwi Power, 2012-13

Power stations that are quick to respond can earn a nice bit of extra money in the Short Term Operating Reserve (STOR) market. All they have to do is ramp up or down by more than 3MW often with as little as 10 minutes warning. Kiwi Power opens up the market for the demand side. They ‘aggregate’ many small loads and offer their combined response capacity to the system operator.

We were given unprecedented access to their client database for research. The findings are one good news and one bad one. The bad: demand responses come mostly from clients with ‘stand by generators’. They do not reduce load at all, but turn on a diesel generator, which makes it look to the outside world as if their load had reduced. This helps with system operation, but not in the cleanest way possible.

The good news: with some changes to the way the market is structured, a lot more genuine ‘turn down’ response is possible. Read more here.

The Role of Electricity Storage in Low Carbon Energy Systems

PhD thesis UKERC studentship (2009-2013), Imperial College London

My PhD combined the techno-economics of storage (does it make commercial sense?) and a socio-technical transition perspective (if storage is desirable to have on the system, how will it enter the market?). Will electricity storage will be a critical part of our energy future.

You want the short answer? Well, two words: it depends. With a high share of wind on the system, storage is not just a ‘nice to have’ technology: my modelling suggests that even at present costs and with modest efficiency it could reduce future system costs significantly (££bn). However, higher shares of CCS and nuclear do rather dampen the outlook for storage. And significant flexibility from the demand side could reduce value substantially (how realistic that prospect is forms part of my ongoing research).
Secondly, the established structures in our systems of energy provision (markets, institutions, culture, infrastructure) do not make it easy for storage to enter the landscape. A pattern emerged when speaking to the broad range of stakeholders: most of them would welcome storage to be part of future systems, but no one feels responsible for investing. This “somebody else’s problem field” surrounding storage may require strategic policy measures, if we are to reap the potential benefits.

A slightly longer answer can be found in my thesis.

The Thread
Storage of Energy

r104FM June 2012. The Thread – live radio thinking. Hosted by Rebecca Wright.
Professor Steven Connor, Dr Jon Agar, Dr Richard Barnett and Philipp Gruenewald talk about energy storage, spanning Victorian fears over ‘running out of energy’ and future needs to save in times of plenty.

Consumer capacity charging

April 2012

The effect of ‘not paying for energy’ on an active demand- side.
Once we get closer to the UK’s targets of low-carbon electricity, the cost structure will have changed dramatically: fuel costs no longer dominate the bill. Instead, over 80% of the cost of electricity is related to the cost of the installed generating capacity (building wind tunnies and power stations – the network is not even included in this figure). Yet, ‘capacity’ is free for consumers, who pay for energy only. This could lead to consumption patterns that require building more capacity than consumers really want to pay for. How much could we reduce peak load on a household level, if the incentive was there? This paper suggests that especially with electrified heating, the scope could be considerable.

Grünewald, P., Consumer capacity charging: the effect of ‘not paying for energy’ on an active demand- side. Energy and People: futures, complexity and challenges conference. Lady Margaret Hall, Oxford 2011.

Hydrogen Challenge 2011

Interdisciplinary team of 10, Imperial College

April 2011

How could hydrogen vehicles be refuelled at home? This is not only a technical challenge – making it economically viable and environmentally sound is also crucial. To tackle this full range of issues a team of 10 Imperial students pulled together for some innovative thinking.

One idea that sprang out of this work is a concept to get better turnover out of the residential refuelling stations. Serving just your own car is poor use of this expensive infrastructure. So shout about it! Broadcast the level of your own hydrogen storage level in real time to a central server. This server informs hydrogen vehicles in your area about available fuel on their sat-nav. If you produced surplus hydrogen (or even run short of storage), your offer price drops and cars will be attracted to fill up at your facility. This way residential refuelling forms part of the national refuelling infrastructure development.

Receiving the award at the H2FC conference in Washington


IChemE award

Innovation and excellence award
Nov 2009
The MSc in Sustainable Energy Futures won IChemE Innovation and Excellence award in the ‘Education and Training’ category, with the judges’ chief criteria including originality and impact on the environment. Celebrating with RWE power CEO Volker Becker at the Houses of Parliament With course director and creator Prof Sandro Maccietto

Winner of the 2009 NPower Graduate Challenge

‘How should a power generation & supply company respond to climate change?’
June 2009
This was the question posed by npower. And if that wasn’t challenge enough, answer it in a 6 minute presentation, please.

So, why not ask npower to change their entire business model?

This is what we did. “CESA” fundamentally changes the way power companies relate to their customers. Instead of charging for the energy delivered, the power company offers to meet all the customers energy needs for a flat fee. Sounds too wacky? Well, there are a few twists to the proposal that impressed the RWE/npower board enough to award us the “npower challenge 2009”.

Celebrating with RWE power CEO Volker Becker at the Houses of Parliament Celebrating with RWE npower CEO Volker Becker at the Houses of Parliament