The term ‘efficiency’ is one that gets used in various contexts within the power system. So it’s worth clearing up exactly what it means when it’s used.
At a basic level, energy efficiency is measuring the energy you get out of a process compared to the energy you put in:
So in a magical world of zero losses, that would be a ratio of 1; which we’d more commonly express as a percentage: 100%.
In power systems, one process for which efficiencies can be stated is that of power (electricity) generation.
The energy output of a power plant is the electricity that it produces, but this electricity doesn’t appear from thin air; it is generated by converting from some primary energy input.
In fossil-fuelled power plants, the primary energy input is the heat produced by burning a fuel. We then convert this heat into electricity using a spinning turbine. That’s why we call them ‘thermal’ power plants.
The heat-producing potential of burning fuel is sometimes described as the energy content of the fuel, the heat of combustion or as its ‘calorific value’. Comparing this fuel energy to the electricity we generate:
So if a coal plant has an efficiency of 40%, the primary energy content of the coal that we burn is a little over twice the amount of electricity produced by converting the resulting heat. The laws of thermodynamics tell us that the other 60% of this heat energy doesn’t simply vanish. In an electricity-only system, this ‘waste’ heat is dumped into the environment somewhere – for example heading into a cooling tower and heating the sky.
To make better use of our primary fuel, we can use that heat for something useful: keeping warm for example. This is a combined heat and power (CHP) system. The electrical efficiency of the power plant may only be 40%. But if we can use as a useful ‘energy output’ the majority of the heat that isn’t converted by the turbine (by pumping it into a district heating system), now the overall system efficiency can be regarded as much higher: 80-90%, for example.
Not every power plant uses a ‘fuel’ as its primary energy source.
Consider a photovoltaic solar panel, which converts solar radiation (light energy) into electricity. The energy input here is the amount of radiation falling onto the panel. When we state that a panel operates at an efficiency of 15%, we mean that of all the light energy falling on it, only 15% of that energy can be captured into an attached circuit as electricity. As in the fossil fuel example, the rest of the energy that entered the panel doesn’t just vanish. It is ‘wasted’ in other ways: passing through, reflecting back or – once again – converting to heat within the panel.
In both fossil fuel and solar power plants, it’s important to bear in mind that efficiency is not a fixed number.
Much like driving your petrol or diesel-powered car, there are more and less efficient ways to burn fuel in a thermal power plant. Turbines don’t like ramping up and down rapidly or operating at speeds (power outputs) away from an optimum ‘sweet spot’. So how a fossil-fuelled power plant is dispatched will determine the long-term efficiency with which it converts primary fuel into electricity output.
In the case of a solar panel, a big influence on conversion efficiency is temperature, with efficiency dropping when a panel is hotter. Ironically, this means that solar panels on sunny days will usually be operating less efficiently than on cloudy days! That’s not just because the ambient temperature is usually higher when it’s sunny, but because the panels are collecting more radiation. Since most of this isn’t converted to electricity, these panels also produce more internal ‘waste’ heat when it’s sunny.
That last example raises an important point that often seems to be misinterpreted. Lower efficiency doesn’t automatically mean lower output. That’s only the case if the energy input is the same.
So just because a solar panel operates less efficiently in full sun doesn’t mean it produces less electricity! On a sunny day there is much more primary solar energy available than on a cloudy day. So this input may convert less efficiently but we still get a significantly higher electricity output, as these example numbers indicate:
Primary energy, efficiency and money
Talking about or presenting data on ‘primary energy’ or ‘primary fuel’ is common with reference to thermal power generation. However we don’t tend to do so when discussing solar (or wind). That’s because a primary fuel is a very different resource to solar irradiation.
Solar energy is present everywhere – albeit not all the time – whether or not we build a power plant to collect it. It is simply an ambient resource that we collect and convert in situ, ideally choosing the sunniest spots to do so. We can’t change how much there is, or produce more to power a new project: the solar resource is a fixed quantity.
By contrast, primary fuels are resources that we have to bring to a power plant, creating a supply chain to do so. We can change how much primary fuel exists by digging or drilling for more. Our primary fuel resources are variable quantities, sized specifically by our need to convert them into useful energy outputs.
Primary fuel needs are therefore useful things to record and monitor, since changes in usage guide strategies to source more (or less) of them, including energy security/dependence, supply chain infrastructure and other important planning considerations.
Charts like the below, a ‘Sankey diagram’, are often used to show country-scale energy flows from primary energy (on the left) to end use (on the right):
At the bottom of the primary energy stack, those thick bands feeding into the middle box marked ‘power station’ represent the energy content of fuels that enter thermal power plants. Whereas those thin bands at the top, for renewable power supply, don’t represent the whole solar irradiation or wind resource, but instead just the electrical output of the power plants using those resources (hence they bypass the ‘power station’ box in the middle).
It’s often confusing that primary energy numbers are added up this way – it would be more logical to exclude those renewables from that red ‘primary energy’ box. It also means care must be taken when considering energy mix data or charts listing thermal along with renewable electricity supply. In too many instances, I’ve seen primary fossil energy (the fuel used or energy input, before conversion) plotted alongside renewable electricity generated (the output). This overestimates the shares of fossil sources in the mix: an apples-to-apples comparison should use the output electricity from all the sources. Take care!
Of course the other point that leaps out from the chart above is just how big that blue ‘lost” box is! While a small amount of this is ‘transmission & distribution’ losses (mentioned below), the majority is as described earlier: much of the energy content in our fuel doesn’t end up as electricity which provides valuable end-use applications (in green). It ends up as waste heat.
The Sankey diagram is thus a stark example of the impact of power generation efficiency on a large scale: much of the energy contained in the fuels we burn (having first sourced and transported them) provides no useful end-use benefit.
It’s even starker when we consider the other difference between primary fuels and primary resources like sunshine or wind: money. While the sun and wind don’t cost anything, the exploration, production and supply chains associated with fossil fuels mean that the latter certainly do.
Every time we burn a unit of fuel we are burning something we’ve paid for. We recoup that investment by selling electricity, so efficiency is a big deal economically: low efficiency means we have less electricity to sell for every unit of fuel we’ve bought. So we either make less money or we have to sell that electricity for a higher price. Or, if the need is to produce a particular amount of electricity over time, a low efficiency plant needs to buy and burn more fuel to do so (higher opex).
That’s not to say efficiency doesn’t matter to a solar plant too. It does. As in the fossil fuel case, low efficiency means less electricity output than from a more efficient power plant of the same size. Unlike the fossil fuel case, if the aim is to produce a particular amount of electricity, a solar power plant can’t buy more sun to compensate for less efficient solar panels. Instead it needs to collect more sun: so it needs more panels, covering a wider land area. It’s more capex spending rather than more opex.
From generator to end-user
Briefly mentioned earlier – and evident on the Sankey diagram – was the issue of transmission and distribution losses.
The electricity that is metered leaving all the various power plants in a system never equals the electricity metered on delivery to all the various end-users in the system. Some of that energy is lost within the grid – if you’ve ever walked beneath buzzing power lines, you can hear some of that energy loss in action.
Given that the grid is very much an energy in / energy out system, we could describe its performance in terms of efficiency. However we rarely do; it’s much more common to talk in terms of the percentage of energy loss, not the percentage retained.
So you’ll read or hear that grid losses are 8%, say, rather than that the grid is 92% efficient.
Efficiency in end-use
Finally, another context in which the term efficiency is commonly used is at the end-use part of the system: in electrical devices such as fridges and TVs.
However the usage of efficiency here is very different – and usually much less precise – than we are used to when discussing power plants. No one talks about their fridge being ‘X’% efficient, they talk about it being A-rated. If they want to buy a more efficient one, to lower their electricity bill, they look out for an AA-rated one instead.
In an electrical device setting, we could regard energy efficiency in the same way as we started off this article: as a ratio between energy out and energy in.
In this case energy in is the electricity we consume, which is easily measured (and paid for!). But what about energy out? In physics terms you’ll often hear energy described in terms of ‘work done’: i.e. an energy conversion that produces some useful outcome.
But how would we measure that for a television? Is it the conversion of electricity to light, for us to watch. Or to the sound that we hear? What about running the graphics chip that presents us with the TV guide? Or the hard disc or memory chips on which we can record programmes?
Clearly it’s not easy to measure in strict energy terms a handy, single ‘output’ number; because there are a variety of energy conversions going on in our device, all of which are useful. In any case, from an economic point of view, the consumer just cares about how quickly their electricity bill will rack up while they watch TV.
So when ratings are given for ‘energy efficiency’ in end-use devices they are relative measures, based on benchmarks. They are focused on the ‘electricity in’ aspect, since that’s what matters to consumers: to keep our food at the same temperature over the same period, will one fridge costs more or less to run than its competitor? The quantification of the ‘energy out’ or ‘work done’ part is too multifaceted to be practicable and has less economic relevance anyway. Whereas a power plant converts an energy input into an energy output which it sells on, a consumer simply pays for an input in order to consume for themselves the outcome of how this energy is used. They aren’t selling their fridge’s cooling or their TV time on to someone else.
The energy chain in summary
The energy journey from our primary energy resource to an eventual end-use application, can be summarised as below. There are various conversions en route – in detail, many more than discussed here! Every time we have a conversion, we can talk about how efficient that conversion is, in energy terms.
In practice, the most common and quantified discussion of efficiency happens around the power generation part. That discussion neatly introduced us to the important concept of ‘primary energy’, including why efficiency matters very differently between thermal and renewable power supply.
Endnote: Efficiency and Capacity Factor
I sometimes encounter students confused by if and how efficiency and capacity factor relate, so it’s worth a few closing comments to clear this up.
Remember that the capacity factor of a power plant is defined as the actual energy output of a power plant during a particular time period divided by the output that the plant could achieve, if it operated at its full rated power for the whole of that same period.
It is in effect a metric which compares how often two different power plants operate close to their rated capacity. It doesn’t say anything about the efficiency of those different power plants.
For example, if the fastest rate that I can throw primary energy at a power plant is 100MW and it has a conversion efficiency of 30%, then the fastest rate it can output electricity – its rated capacity – is 30MW. If that plant operated for an hour at full capacity, it would generate 30MWh of energy.
If the conversion efficiency of a second power plant is 40% then, for the same rate of primary energy input, its rated capacity will be 40MW. It should be apparent now that efficiency is already baked into a power plant’s capacity rating: throw the same primary energy at two plants and the more efficient one will have a higher capacity. For our second plant, an hour’s full-capacity operation would yield 40MWh.
If both plants actually produced 20MWh over an hour, the least efficient one would have a capacity factor of 20MWh/30MWh = 66%. By contrast the more efficient one would actually have a lower capacity factor, of only 50% (20MWh/40MWh).
We know the efficiencies of these two power plants, so we can work backwards and calculate that the least efficient one used an actual primary energy input of 66.6MWh over the measured hour (= 20MWh / 30%), while the more efficient one used only 50MWh (= 20MWh / 40%).
In other words the lower capacity factor is actually a result of a lower primary energy input: a thermal plant choosing to burn less fuel or a solar farm sitting on a less sunny patch of land. (Use the same primary energy input for both our example plants and the capacity factors will be the same too – try this calculation for yourself!).
In real power systems, efficiency can certainly influence the capacity factors of thermal plants, because it influences that choice of how much fuel to burn. That’s because of economics: fuel costs money, so power plants which make better use of it are likely to be dispatched more often. Inefficient plants, particularly ones burning expensive fuel, are used more rarely; as ‘peaker’ plants for example, with very low capacity factors.