We’ve seen that peak demand periods place requirements on the electricity infrastructure that are disproportionate to their duration and frequency. Since generation, transmission and distribution capacity must be maintained and operated to meet these demands, the economic impacts are significant too. Those generators which need to turn on for just a few hours per year are the ones which are most expensive to run, meaning that peak power is expensive; so extreme spikes in price can occur.
Rather than keep building more and more power capacity, more and more power system operators and policymakers are concluding that a better economic solution is through some form of “demand management” or “demand response”.
Traditionally, the prices that consumers of electricity pay varied relatively little. Simple differences between day and night tariffs, for example, didn’t capture either the variability in demand during these periods or the variations in the cost of generating electricity at different times (as different supply options turned on or off). In a system where the electricity demand and generating cost at, say, 11am might be considerably different to that at 3pm, if there is little or no difference in tariff to capture that difference, there is little incentive for consumers to alter their behaviour. In particular there is no incentive for them to reduce their consumption when demand is high, which would be helpful for the overall system, because they are spending no more than when demand is lower.
In other words, power systems have traditionally allowed demand to vary with few economic constraints; and then relied on the adjustment of sufficient capacity of flexible supply to be able to meet it.
Increasingly though, system operators and policymakers are encouraging electricity consumers to reduce their consumption when demand is high.
If a peak demand period can be reduced or removed, it can save on capacity requirements, particularly those least used, highest cost ones. This provides a much more economic and sustainable approach in the long-term, and one better able to incorporate more variable generators (limiting demand when there is low wind availability for example).
Usually this “encouragement” involves a financial benefit for consumers . This could be the ability to gain a better (cheaper) overall tariff if they agree to reduce their electricity usage at particular times or in reaction to specific signals from the system. Or it can involve being specifically paid to reduce demand when the system requests them to do so. While it might seem odd to pay electricity consumers not to use electricity, bear in mind that the alternative – to supply them by turning on new generation for just a short time – can prove much more expensive.
While on an individual basis, a single electricity customer may be too small to significantly impact system-wide demand on its own, some markets have seen the emergence of aggregators (1). These are companies who sit between multiple electricity consumers and the system operator. They are hence able to co-ordinate demand reduction in response to signals from the system operator, by adding up many small reductions in demand into the required overall size (in return, of course, for financial reward).
If incentives or market mechanisms fail to change behaviours, then policymakers could choose to require (rather than request) reductions in demand at peak times. Or, in extreme circumstances, simply cut off supply (for example where supply/demand imbalances could lead to blackouts and all the economic damage and grid restart issues that these create).
The subject of demand management is a broad one and includes a fast-changing, fast-growing and complex world of policies, incentives, market mechanisms and more.
However in terms of considering its impact on overall electricity generation supply, you can contrast two simple scenarios.
If demand at peak time is reduced, by means of that consumption simply stopping or being prevented, then the result is also a reduction in overall energy use.
In the chart above, the y-axis is GW and the x-axis covers a period of one day (24 hours, midnight to midnight). The red line is the original power demand curve during the day, peaking at 24 GW at 4pm. Remember that the energy used is the area under this red line (power on the y-axis multiplied by time on the x): it adds up to 407 GWh. The difference between peak and minimum/baseload demand during the day is 14 GW
The green area represents the energy used when demand is allowed no higher than 20 GW. It adds up to 393 GWh.
So the amount of energy saved by “shaving” off the top 4 GW of demand is the white area between the red curve and the top of the green area. The difference between peak and minimum demand during the day is now 10 GW.
Perhaps more likely is that when consumption is reduced at peak times, that same consumption will still be needed at some point (to do whatever work the user needs to do). It might simply shift to a time of lower demand (and price), for example overnight.
In the chart above, the red curve is the same as in the previous chart, and the green area once again represents a scenario where peak demand hasn’t been allowed to rise above 20 GW..
However in this case, the energy consumption removed from peak demand periods has now been added as additional (green) demand overnight – the consumption hasn’t disappeared, but has been time-shifted. The green areas above the red line are exactly the same as the white areas below the red line. Energy demand over the day is 407 GWh, identical in both the red and green cases.
The difference between peak and minimum demand during the day is now 8 GW.
There is of course another way to reduce peak demand. That is to reduce all demand – by means of making consumers and the devices they use (lightbulbs or air-conditioners for example) more efficient:
Starting with the same red demand curve, to reduce peak demand down from 24 to 20 GW requires an efficiency improvement of 16.7%. If this efficiency gain is assumed to apply regardless of time of day, all consumption is lowered – as per the green curve above. Over the whole day, energy consumption is down from 407 GWh to 339 GWh.
The difference between peak and minimum demand during the day is now 11.7 GW; of the three, this solution has had the least impact in lessening the difference between minimum and peak power demand but the most in terms of energy saving.
Of course the reality is that the reduction of peak demand in a system could be a result of a combination of all the three basic scenarios charted above.
Energy efficiency should be a key goal of policymakers (and consumers) aside from any concerns about peak demands. Encouraging consumers to shift their electricity usage from peak to off-peak times makes sense as an additional or parallel measure. If these methods can’t achieve what’s required, then perhaps simply disallowing energy usage at peak times becomes the next step.
(1) This is a good – though terminology-packed – short whitepaper that discusses the role of aggregation in markets: http://smartenergydemand.eu/wp-content/uploads/201…; in particular how to increase it in Europe.
This line from the introductory paragraphs summarises the theme: “Demand Response programs provide consumers (residential, commercial or industrial) with control signals and/or financial incentives to lower or adjust their consumption at strategic times. In market such as the USA, consumers currently (=2015) earn over 2.2 billion Euros annually from these programs, and save more than 5 times this amount for all consumers by avoiding expensive generation capacities and therefore reducing costs for all.
(2) Having been asked once about the impact of electric vehicles (EVs) on peak demand, I wrote this article which explores how significant that might be and – in particular – how a smart approach to demand management is going to be crucial to scaling EV uptake: https://greycellsenergy.com/articles-analysis/smar…