National Grid and Ernst & Young concluded that:

They petitioned for the following policy measures:

And they highlighted that the purpose of the report was to persuade the government to reduce its expectations of other technologies:

There was nothing subtle about the rent-seeking. Nor was there much credible analysis underpinning it.

The core of their claims was an analysis of the resource, which claimed much larger numbers than previous research. The biggest factor in this was the inclusion of gasification gas as well as digester gas, as this allowed them to count feedstocks that could not realistically be converted to gas by AD. In the table below, the last two items were intended for gasification. But they also upgraded significantly the claimed potential for biogas.

How does this compare with reality?

At the time, approximately 180 ktoe (2.1 TWh) of gas was being produced by sewage digestion for electricity generation, and another 50 ktoe (580 GWh) of sewage gas was being produced for heat.[1] That is around 247 million m³.

Waste-water processing needs a considerable amount of heat and power. CHP is a good fit, as much of the energy can be used on site. None of the above gas was being scrubbed and injected into the grid (even assuming there was a grid connection at the sewage works to inject into). There was no reason for the water companies to change their process, so that they would export their scrubbed biogas and import some other forms of energy to run their processes.

Water companies had had practical reasons to produce their own energy for a long time, and had additionally been incentivised to do so under the NFFO and the RO. Consequently, they started from a relatively high level and had already expanded their production considerably by 2009.

Gas production is only practical at the larger sewage works, most of which were already doing so. The question of scale would be doubly significant for grid injection, as the smaller, more remote sewage works were less likely to be on the gas grid, as well as being less suitable for gas production. And to the extent that there was some remaining potential, new sites would have the same reasons as existing sites to deploy CHP rather than grid injection.

By 2018, sewage gas production had increased to 325 ktoe (3.8 TWh) for electricity and 83 ktoe (965 GWh) for heat. That is equivalent to around 438m m³. Despite the incentives being skewed to favour biomethane, almost none of it was being injected into the grid.[2]

There was never much prospect that 270m m³ would be diverted from CHP to grid injection. There was even less prospect that 629m m³ of sewage gas would be produced by 2020, let alone be diverted 100% to the grid.

One of the first commercial AD plants in the UK, Holsworthy Biogas, was originally obliged to take 80% of its feedstock from farm slurry. It was not economic under this constraint, and went into administration. It only became viable when its new owners (Summerleaze) broke the obligation and switched the plant to mostly food waste.

Cattle manure and slurry produce a fraction of the gas per tonne of (wet) feedstock, compared to food waste, and less still than energy crops. Waste attracts a gate fee (albeit significantly diminished nowadays due to over-capacity), whereas farmers will not generally pay to get rid of their manure. The manure intake therefore represents a cost not an income, because it has to be transported to the digester.

On-farm digesters may mitigate some of these disadvantages by minimising the transport distance and utilising the energy on-site, which offers a higher value than export as gas or electricity. Specialist farm digesters can also be cheaper, as they can be simpler than digesters that handle waste. And digestate disposal directly on the farm may be cheaper.[1]However, they will tend to be smaller, and will often not be located on the gas grid. In the rare cases where the benefits of localism are sufficient to outweigh the poor economic fundamentals of slurry digestion, the opportunities for grid injection may be limited.

There is thought to be a substantial resource of manure and farm slurry. In England and Wales, it is estimated that approximately 72m tonnes is collected for spreading, and another 73m tonnes is deposited directly in the fields by the animals.[2] Of this, dairy and beef accounts for nearly 60m tonnes of the spread material and 42m tonnes of the directly-deposited material. At around 15 m³/tonne, this implies a theoretical potential of 900m m³ of methane from AD, if all the collected dairy and beef manure/slurry were digested before spreading.

Yet the production of biogas from this material was negligible at the time of the NG/E&Y report, and remains insignificant today, despite the stimulus efforts. A 2017 report by Ricardo for ClimateXChange identified:[3]

Bioenergy Europe note that:[4]

Yet biogas as a whole constitutes 2.1%, 0.4% and 1.3% of those countries’ primary energy consumption. Agricultural residues constitute 51%, 68% and 49% (by mass) of their biogas feedstock. Given the lower gas-producing potential of the feedstock, that means that in the European countries that have placed greatest emphasis on manure as a biogas feedstock, it is responsible for around 0.75%, 0.2% and 0.4% of their primary energy consumption. That is particularly striking in Denmark, which is noted for its successful promotion of renewables, has the infrastructure (e.g. municipal district heating schemes) to maximise the value of AD plants, and has large quantities of pig (and other) slurries to digest.

If we combine the fact that farm slurries are generally uneconomic but may have limited potential as a complement to higher-gas feedstocks, with the experience in the countries that have tried hardest to use this material, we may conclude that the maximum practical proportion of the feedstock is around 50% by mass, which equates to around 25% of the gas potential (i.e. if we work out the potential of other feedstocks, we could add up to 33% for co-digestion of this material).

The reality of digesting manure is that, while the theoretical potential is quite large, the economic potential is small and always has been. NG/E&Y were not making abstract claims about the theoretical potential. They were claiming not only that biomethane could supply nearly half our domestic gas, but that the level of support needed to achieve that was modest. The combination of these two claims with regard to manure is disingenuous.

Agricultural waste was brought within waste management regulations in 2006. The vast majority of agricultural waste was the slurries and manures. Of the remainder, the focus was on handling the plastics.[1]

There was growing consideration of anaerobic digestion, and more limited consideration of on-farm non-slurry agricultural waste, but little consideration of the two together, perhaps reflecting the reality that the quantities were modest and already used for practical purposes (e.g. animal feed or composting).

In the Anaerobic Digestate protocol produced by WRAP in Feb 2009, for instance, agricultural waste was treated as though it was entirely about slurries and manure.[2]

The 2007 report by the Biomass Task Force (chaired by a former NFU chairman) focused strongly on the role of agriculture, but did not identify any “wet” agricultural wastes other than slurries suitable for digestion.[3] It did highlight significant volumes of “dry” agricultural waste such as straw, suitable for thermal processes (e.g. gasification).

The Environment Agency’s recommendations for Agricultural Waste, published in 2001 offered the following figures:[4]

It is likely that a lot of the animal material would have needed to go to rendering. 100m m³ of biomethane looks like a stretch as the potential of this resource if it were all digested. And there was no reason to think that it would all be digested, rather than continuing to be used for the valuable and environmental uses (e.g. animal feed) to which it was being put.

It has not been possible to identify any quantification, nor much reference, to non-slurry agricultural wastes being used as AD feedstocks at the time of E&Y’s report. A small number of on-farm digesters fed primarily with slurry/manure were noted, and it is reasonable to infer that they would also have use any other putrescible material available. But it is hard to see on what basis a researcher could have concluded that there was any significant current contribution from this material in 2009, from which to project a large contribution in 2020.

Yet E&Y estimated the baseline contribution at 234m m³, and the “stretch” contribution at 967m m³ – nearly double the “stretch” contribution of the slurries and manures that dominated the statistics and analysis of the resource.

Fast forward to their target date, and WRAP recently estimated the amount of on-farm food waste and surplus produce as 3.6m tonnes.[5] Their objective, as always, was to minimise this waste, not to maximise its use for digestion. Most of it is already re-purposed, for example as animal feed. As the 2017 report for ClimateXChange, cited above, notes, there are few known instances of the digestion of this material to produce gas, whether for on-site use or grid injection. There is little reason to think now that a significant amount of gas will be produced from this material as efforts to minimise it proceed, and there was little reason in 2009 to expect much either.

Other than sewage, food waste (post-farm-gate) is the main feedstock (around 50%) of anaerobic digestion in the UK, because it has a good gas yield and historically some waste-disposal income (i.e. gate fee).[1]

Image

There was very little anaerobic digestion (other than sewage and landfill gas) in 2009. This was one area where there was genuine potential for a material increase in gas production, whether for electricity, heat or biomethane. If policies had directed all the material to biomethane, E&Y could reasonably have claimed all the gas potential from this material.

Food waste was a hot topic in the years up to 2009. On the upside, the resource was large. 3.6m tonnes of domestic food was thought to be wasted annually.[2] Estimates of Commercial & Industrial (C&I) food wastes varied widely, depending on reporting and definition, but pointed to around 3.5m tonnes excluding compostable/garden waste.[3]

On the downside, the policy preference for this material was not to utilise it for energy, but to reduce it.[4] That, of course, would reduce its potential as a feedstock for AD. Half the resource (the domestic element) was not available to digesters unless councils chose to implement separate collection, and would depend on the efficiency with which the public separated their waste.

With full source-separation and collection, and no waste-reduction, this material had the potential for up to around 750m m³. A prudent assessment would have assumed some minimisation and incomplete collection, and some use for the generation of electricity and/or heat (e.g. where large food-waste producers produced their own gas for their own processes on site).

History bares out the prudent judgment. Significantly more of the resource went to electricity and heat than to biomethane.[5]Separate collection remains incomplete. Significant efforts are being made to reduce the volume of food waste, or to divert the material to other uses perceived as higher value.

WRAP are actively promoting the diversion of material that currently goes to anaerobic digestion to uses such as animal feed.[6]

WRAP announced recently that food-waste volumes fell 7% between 2015 and 2018, an acceleration of the trend from 2007-15, when volumes fell around 11%.[7] They nevertheless note that a significant proportion remains uncollected and the public continue to under-estimate the significance. The Committee on Climate Change’s recent report on and plan for land-use envisages another 20% reduction by 2030.[8]

Reality is demonstrating what could reasonably have been inferred by a knowledgeable analyst in 2009: there was plenty of scope to increase the digestion of food waste, but it was a declining resource that was unlikely to be captured fully for the purposes of generating biomethane. A range of 250 – 750m m³ represents an optimistic assumption that one-third to 100% of it would be captured and used for biomethane, and the decline would not be significant. It is hard to understand how one credibly arrives at a higher figure than that.

The greatest mystery about NG/E&Y’s figures for biodegradable waste is exactly what material they had in mind, and how they expected it to be used. The main biodegradable components of municipal and commercial wastes are food wastes. But these are considered separately.[1]

Other technically-biodegradable components of the waste stream include garden waste, paper, card and wood. None of these is very suitable for digestion. Wood waste is also considered separately, but for wont of an alternative explanation, the rest must be the core of what NG/E&Y had in mind for this category.

Garden waste was and continues to be composted primarily. There is an established infrastructure and market for this product. It is unsuitable for digestion, because of the amount of fibre/lignin/non-putrescibles. It could be suitable for gasification (see below) if suitably prepared, but it is doubtful that the challenge is worth the reward compared to the straightforward and commercial option of composting. Were it diverted from composting, then there would be a danger of the market resorting to less environmentally-friendly options such as peat.

Paper and card are also not suitable for digestion, at least without significant technological breakthroughs that were not in prospect in 2009 and have not materialised since. They are eminently suitable for thermal processes, and form (along with plastics) the bedrock of the UK’s significant expansion of Energy-from-Waste. They could be gasified instead of incinerated, if the technology were sufficiently mature and competitive (see below). Much of it can also be (and substantially is) recycled.

The recycling rate for paper and cardboard is widely reported to be around 80%.[2] Government statistics provide only partial information, and are structured in such a way that it is difficult to get a complete picture.[3] The real picture appears to be around 7.5m tonnes of paper recovered from around 10.8m tonnes of paper and cardboard consumed.[4]The unrecoverable proportion is thought to be around 22%, which leaves around 970,000 tonnes available for gasification, if it can be recovered and converted to a useful fuel economically, without diverting material from currently-preferred recovery options.

The recycling rate for paper and card was already high by 2009.[5] And total volumes were higher, as the cyclical and structural declines in paper consumption had only just begun. However, over half of the recovered material was exported, mainly to China. There were problems with demand from both native and foreign recyclers.[6] NG/E&Y might reasonably have assumed that most of the exported, some of the native-recovered and some of the unrecovered material could be available for gasification – perhaps as much as 5-6m tonnes. This would have turned out to be much too optimistic, but the availability of feedstock is anyway one of the lesser problems with this option to increase the volumes of green gas (see below).

Of the waste streams listed by NG/E&Y, at least two – wood waste and miscanthus – must have been envisaged as feedstock primarily for gasification, not digestion. As explained above, we should probably add (despite its name) the biodegradable waste, as effectively the paper and card waste stream.

Of approximately 4.5m tonnes of waste wood produced in the UK, around 2.1m tonnes is currently going to biomass energy projects.[1] 1.35m tonnes is reused and 0.3m tonnes is exported.

The total amount of waste wood was not dissimilar in 2009.[2] The majority was from construction and demolition waste, which introduces challenges of identifying and handling the treated wood. WRAP estimated that only around 1.4m tonnes was “clean solid wood”.

WRAP seem to indicate that almost all of it was already being recovered, driven primarily by the wood panel industry.[3] Nevertheless, WRAP expected recovery to increase overall, and energy-use to increase (driven by the RO, i.e. for electricity generation) while panel manufacturers’ share fell.[4]

Barring the complete implosion of British panelboard manufacturing, it was reasonable for NG/E&Y to assume that a material proportion of the clean waste wood would continue to go to this use. Likewise, animal bedding and other existing forms of re-use.

Around 2m tonnes of waste wood available for energy-use, seems to be a recurring figure from the studies, which would have been a reasonable top-end figure for gasification assuming it displaced all other energy-uses. It is also consistent with what happened in practice under the incentives of the RO and RHI, although of course it went almost entirely to mature combustion technologies and not to gasification.

We set out below what might be credible if the UK made ambitious efforts to produce energy crops for digestion. The constraining figures are the same for gasification feedstocks. The same amount of land is available, and the uses are mutually exclusive.

A key difference is that the technology for digestion is mature. We might therefore prioritise energy-crop land-use for digestion rather than gasification. Nevertheless, not all land is equally suitable for all crops, so there might well be a mix. Some proportion of the figure estimated below for energy crops might have been available for gasification instead of digestion. In practice, 7,000 ha produced around 71,000 odt of miscanthus in 2018.[5]

So we have a maximum potential gasification feedstock of around 5-6 million tonnes of paper and cardboard, 2 million tonnes of waste wood, and perhaps 6 million tonnes of energy crops.[6] That’s around 63 TWh of feedstock (which does not mean 63 TWh of potential biomethane, because of conversion efficiencies, as below). Around 27 TWh of that is mutually exclusive with the upper estimate for digestion energy crops.

The true problem for gasification starts when we consider how this would be used. Gasification technology is in one sense very mature. It has been known and used for decades.[7] But in another sense, it is still immature. Despite many attempts in recent decades to commercialise it, no offering has succeeded to any significant extent.

One problem relevant to this report is the nature of the feedstock. Gasification was only ever deployed at scale for the conversion of coal, which is a homogeneous fuel. Recent efforts have focused on more heterogeneous feedstocks, such as waste. But the process needs very precise control of energy-feed-rate, air, etc. Many gasification projects have foundered on the front-end.

Another major problem for these purposes is that gasification does not produce biomethane. It produces a mixture of methane (CH₄), carbon monoxide (CO) and hydrogen (H₂), to name the three main combustible elements of the gaseous product. In the days of town gas, this mixture was permitted in the gas network. Nowadays, there is a very tight specification for the gas that can be accepted into the network: largely methane with a small amount of more-complex alkanes. For operational and safety reasons, gasification gas could no longer be added to the gas network. So one immature technology (gasification) would have to be complemented by an even less mature technology (methanation) to convert the gasification gas into a gas that could be distributed to users.

Considerable effort has been invested in recent years in gasification + methanation to produce a Synthetic Natural Gas (SNG). But (a) it is 2020 and none of those are commercially mature, and (b) NG/E&Y could not reasonably have predicted that they would be, given the state of technology and the market in 2009. Decades of efforts around the world had resulted in no commercial options to date. The prudent assumption was that 10 years was not long enough for the problems to be ironed out and the technology commercialised.

Yet NG/E&Y included the following quantities of biogas from gasification in their baseline projection: 1,253m m³ from wood waste, 1,845m m³ from miscanthus, and (probably) 1,042m m³ from biodegradable waste. This constituted 74% of the 5,625m m³ of biogas that they envisaged would be available as a minimum in 2020.

Of all their imprecisions, this is the most egregious. The only reasonable figure to use for this technology in the baseline scenario was zero. Decades of effort had so far delivered almost nothing. And sure enough, from the perspective of 2020, we find that the most recent decade also produced almost no commercial SNG production, not just in the UK, but globally.

As for the “stretch” scenario, how much biogas might those 63 TWh of feedstock have produced? The conversion efficiencies of gasification and methanation are both typically estimated at around 80%, so we will assume the combined efficiency would be 64%. That implies a maximum of around 40 TWh of biomethane from this source.

At 10.83 kWh/m³ of methane, that’s around 3.7bn m³ of biomethane. NG/E&Y had 2,697m m³ from wood waste, 3,971m m³ from miscanthus, and 8,328m m³ from biodegradable waste (probably). This technology now accounted for 81% of the total 18,432m m³ of biomethane anticipated in the “stretch” scenario.

In other words, gasification and these feedstocks, were NG/E&Y’s deus ex machina to convert what was obviously an inadequate potential of digester gas, to an apparently-significant potential of biomethane. Unfortunately, it was not just a misestimation; it was a fiction with no basis for reasonable expectation given the knowledge available at the time.

Ernst & Young considered only “a possible limited contribution from sustainable energy crops” for gasification (none for digestion), even though Enviros had relied heavily on it in their most ambitious scenario for the potential of biogas.^[1] The “food vs fuel” debate was a hot topic, and E&Y no doubt judged that it would raise political resistance to rely on energy crops.^[2]

In practice (and in theory to most analysts, then and now, other than E&Y), energy crops are the only way to expand biogas production beyond the limited potential of waste resources. But there is a practical limit even to the grandest ambitions.

Image

Germany has made by far the largest effort to develop biogas.^[3] It has relied heavily on energy crops to do so.^[4]

Germany has devoted 14% of its agricultural land (2.35m ha, 8% of its undeveloped land) to energy crops. Of this, nearly 60% (1.37m ha) is for biogas feedstocks. Yet biogas represents 2.2% of Germany’s primary energy consumption.^[5] Biogas represents around 5% of Germany’s electricity and around 1.4% of its heat, and an insignificant proportion of their transport fuels.^[6]

Germany illustrates the competition for land where significant reliance is placed on energy crops. Most of the rest of the energy-crop land (964,000 ha) is for the production of liquid fuels (biodiesel and bioethanol). Liquid fuels are hard to replace in some uses, such as aviation fuel. Biofuels are not currently used much for aviation in Germany (or elsewhere). They represent 4.7% of the fuel consumption for land transport.^[7] If used for aviation, this volume would represent around 24% of Germany’s requirements.^[8]

Nearly one-third of Germany’s land is forested, compared with 13% of the UK’s land. The Germans have utilised solid biomass (primarily wood) to a greater extent than the UK. The Germans do not waste so much of this resource as the UK generating electricity at around 35% conversion efficiency. Most of it is used to produce heat (direct or as CHP) at around 80% efficiency.^[9]

Image

Image

The UK has a lot less land than Germany dedicated to the production of solid fuels (e.g. wood), liquid fuels (biodiesel/ethanol) and gaseous fuels (biogas). If we focused on expanding one aspect (e.g. energy crops for biogas), we would miss the contributions of the other types of energy. And all of them compete with the traditional use of agricultural land for food production.

Enviros envisaged the use of 157,000 ha for the production of energy crops for AD in their most ambitious scenario. Ernst & Young discounted the possibility altogether.

94,000 ha are used to grow energy crops in the UK, out of a total utilised agricultural area of 17.5m ha.^[13] We have 3.19m ha of woodland.^[14] Our total land area is 24.4m ha.

Germany has 2.35m ha growing energy crops (out of 16.7m ha of agricultural land) and 11.4m ha of forest, out of a total land area of 35.7m ha. Yet Germany meets only a small proportion of its energy requirements from these fuels. Is there (and was there ever) a credible model for the UK to produce a material proportion of its energy requirements from a smaller area of land? And if not, what is the model for biogas to make a material contribution to the UK’s energy supplies, given that it can only be stretched to Germany’s modest contribution through orders of magnitude higher use of land for energy production than we currently deploy?

Let’s imagine that the UK increased the level of energy crop production to the same as Germany. 14% of UK agricultural land = 2.45m ha. 58% of this (1.43m ha) is used to produce AD feedstocks, while 41% (1m ha) is used to produce liquid biofuels. The German experience suggests around 45MWh/ha p.a. for biogas from energy crops.^[15] We could produce 64.35 TWh of biogas, i.e. around 5.9bn m³. That represents around 17% of domestic gas consumption or 6% of total gas demand, according to Ernst & Young’s projections. In practice, gas consumption has fallen somewhat, and this amount would represent 21% of domestic gas consumption and 7% of total gas demand.

We need additionally to allow for the reforestation that is (a) part of this government’s policy, and (b) required to match the German model. To bring us up to German levels of afforestation, we will need to switch another 4.6m ha from agricultural use to forestry. Our agricultural land area would be reduced from 17.5m ha to 10.5m ha.

Or we might choose to have less forest than Germany. In that case, we will either have to import more solid biomass fuels, or use less solid-biomass heat, which currently makes up approximately two-thirds of Germany’s renewable heat.

For reasons of climate policy, the UK (like many countries around the world) plans to increase significantly its forested area. Without deforestation or a substantial reduction in agricultural production, the potential for energy crops in the UK will have to be kept within reasonable bounds that constrain the potential of digestion or other conversion technologies.

---

[1] See above: 157,000 ha of land for energy-crop production, contributing to total biogas production equal to <4% of total heat demand.

[2] e.g. David J Tenenbaum, “Food vs. Fuel: Diversion of Crops Could Cause More Hunger”, Environ Health Prospect, Jun 2008. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2430252/

[3] Bioenergy Europe, Statistical Report 2019, Biogas section Table 2

[4] Fachagentur Nachwachsende Rohstoffe, Bioenergy in Germany, Facts & Figures 2019, http://www.fnr.de/fileadmin/allgemein/pdf/broschueren/broschuere_basisdaten_bioenergie_2018_engl_web_neu.pdf

Image

Image

Given the different gas potential of energy crops and excrements, this means that energy crops are responsible for around two-thirds of Germany’s biogas.

[5] Bioenergy in Germany, Facts & Figures, p.39: used share of biogas primary energy potential: 273 PJ = 75.8 TWh. Eurostat, T2020_33, Primary energy consumption. Germany (2017): 298.31 ktoe = 3,469 TWh https://ec.europa.eu/eurostat/databrowser/view/t2020_33/default/table?lang=en

[6] Data from Bioenergy in Germany, Facts & Figures 2019. Gross electricity generation: 654.8 TWh. Electricity generation from biogas: 63.2% of 51.4 = 32.5 TWh. Renewable heat (162.2 TWh) share of final energy consumption for heat: 12.9%. Biogas share of renewable heat: 10.6%

[7] ibid. p.28. Total land transport fuel consumption: 57.6 Mtoe = 670 TWh. ∴ Biofuels = 31.5 TWh

[8] Germany’s jet fuel consumption 2018: 221,010 barrels of oil per day = 131.3 TWh p.a. https://www.theglobaleconomy.com/Germany/jet_fuel_consumption/

[9] ibid. pp.5-6.

[10] Excludes biogenic fraction of waste. UK figure based on DUKES 6.1.1 depends on highly-suspect estimates of domestic wood burning. Reality is probably roughly half.

[11] Excludes sewage gas and landfill gas

[12] UK: 599m litres @ 9.8 kWh/litre. Germany: 4.7% of 57.6m toe total land transport fuel consumption.

[13] 27,000 ha are for liquid biofuels. 57,000 ha are for AD (primarily maize). 10,000 ha are for solid biomass. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/856695/nonfood-statsnotice2018-08jan20.pdf
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/747210/structure-jun2018prov-UK-11oct18.pdf

[14] https://www.forestresearch.gov.uk/tools-and-resources/statistics/statistics-by-topic/woodland-statistics/

[15] This assumes mostly the highest-yielding crop – maize silage – is grown, but some land is not suitable and requires lower-yielding crops.