The International Energy Agency (IEA) “under-reports (the) contribution solar and wind by a factor of three” compared to fossil fuels, according to a recent report. I and others have been pointing this out regularly, but it’s good to see this methodological anomaly (if that’s what it is) exposed and explored in more detail in an article by Erik Sauar.

His rendition of it is a little complex and convoluted in places, but the basic point is that the IEA makes use of primary energy data, which, for fossil fuels, are relatively straight forward: the tonnes of fuel used by power plants, usually rendered as million tonnes of oil (mtoe) to make it comparable in energy content terms. However, for renewables, since you can’t really measure the input raw energy, for wind and solar they just use the output energy to represent the primary mtoe figure. But strictly, to make sensible comparisons (e.g. of the relative carbon emission implications), a way to infer the primary energy for these renewables is needed. One way is to work out the amount of fossil fuel that would have to be used to produce the same energy output. Standard steam-raising fossil plants are very inefficient, typically wasting two thirds of the primary energy fed in to them, rejected as heat into the environment. So you would need about three times more fossil energy input to get the same output as from a similarly rated renewable generator. Hence Sauar’s claim that some renewables are in effect de-rated unfairly by around a third.

Sauar notes that something like the “grossing up” he thinks is necessary is actually done in the case of biomass and nuclear. He says the IEA has chosen “to only measure the electricity produced for these two energy sources, and thereafter multiply it with 3.0 for biomass and 3.03 for nuclear energy”, to get a primary energy figure. That may make sense, even though for biomass, its output is already subject to the aforementioned losses when burnt for power production. But presumably some adjustment would be needed to take account of the lower energy content/tonne compared with oil. Sauar, however, does not change the figure for biomass in his correction of the IEA data, and he also leaves the IEA nuclear primary energy unchanged, just adjusting wind and solar. However, his take on wind and solar seems slightly odd – he says solar needs a 5 times mark-up, wind just 2.3 times, since he includes losses in their conversion processes. That’s debatable. Their “internal” technical efficiencies (which he puts at 20% and 50% respectively) are surely a separate issue, and in any case their load factors vary by location and over time. It seems odd to try to track back to some pre-conversion stage “virtual” renewable energy input and present that as the primary energy.

It makes more sense to work on the final actual output, however derived, suitably grossed up to mtoe, as if it was delivered by a fossil plant, if you want a primary energy figure. He says that BP now does that and uses a 2.8× conversion figure. If you look at the company’s recent statistical review, it says the data for hydro, wind, geothermal, solar, biomass and waste, and also nuclear, are “converted on the basis of thermal equivalence assuming 38% conversion efficiency in a modern thermal power station”. So it seems BP is grossing up renewables properly, including biomass, and also nuclear.

Even so, there’s still room for methodological disputes. Indeed, evidently in response to Sauar’s article, the IEA produced an analysis, which noted that “The IEA had at a point used the ‘partial substitution method’, based on the assumption that hydro, wind, solar electricity had displaced thermal generation. This involved using an average thermal conversion efficiency (e.g. 36%) to back-compute their corresponding ‘primary energy equivalent’. This made their shares in the primary energy supply greater (around three times as much). However, the principle was abandoned as it relied on arbitrary conversion factors and was creating some transformation losses inside the energy balance that did not really exist.”

Nevertheless, their version of the primary energy data is still problematic. Maybe it’s best to use output data. That’s what Michael Liebreich from Bloomberg New Energy Finance said in response to Sauar’s article: “Using primary energy to compare the contribution of different energy sources, rather than final energy, needs to be consigned to the dustbin. It rewards the least efficient technologies, making them seem irreplaceable, at the expense of modern renewables.”

The debate goes on, although there does seem to be some confusion and the need for a standardized approach. See Energy Matters’ take on it, based on their uncertainties about the EU’s handling of data.

We do need to sort this all out. Does “grossing up” really make sense, especially for steam-raising nuclear? Does that give us a useful number?  It would get even harder with hybrid systems: how would hybrid solar thermal PVT fare on this analysis? And nuclear, biomass or geothermal combined heat and power? Which bit of the output would be grossed up and which bit left alone? You can see why final megawatt-hour output is maybe a more solid metric. It tells us what is actually available, always assuming the data have been collected live and not just inferred from the megawatt capacity and assumptions about load factors and likely plant availability.

There are plenty of other potential energy data issues that also need addressing. For example, it’s vital that a clear distinction is made between energy and electricity. Too often, in media presentations and even company PR, contributions are cited as if they are a percentage of “energy” when it’s actually only of electricity. For example, it is sometimes claimed that the Hinkley, Moorside or some other nuclear plant will supply “7% of UK energy”, when in fact it’s only 7% of electricity – the energy contribution would be more like 2%, depending on how you do the sums: are we talking about ex-plant output, or (somewhat less, after transmission) final electricity use?  This type of error is regularly pointed out.

It doesn’t help that the term “power” is often used to mean electricity (and we talk of electric power, power outputs, power plants, and even combined heat and power) when what is meant is electricity. In any case “power” and “energy” are different things: power is the generation or consumption capacity of a device, measured in watts (or multiples of watts), the energy they then supply or use is a time based-measure, calculated in watt-hours (and multiples). So a 1 kW power-rated electric fire run for an hour uses 1 kWh of electrical energy, while a 1 MW-rated wind turbine, if the wind was such that it could run at its full power rating for an hour, would generate 1 MWh of electrical energy. Simple really, but even the best of us occasionally use “power” as shorthand for electricity.

And here’s another potential screw up: Blockchain and all that

The Bitcoin e-banking system uses block chain electronic encryption exchanges to transfer credits and its booming. The same idea is now being touted for wider use.

It has issues and then some – it’s very energy intensive. Also see this analysis and this review. There are maybe better ideas. Though is this really sensible as a fix for it?