An alternative method for estimating the potential

There is a simple way of estimating the gas-producing potential of a large proportion of the wastes under consideration, which was available and almost as accurate in 2009 as it is now.

Most of our putrescible waste has been going to landfill for decades. For around 30 years, they have been engineered to contain the gas (so far as possible), regulations have required the control (i.e. extraction) of the gas, and incentives have encouraged the collection of the gas as an energy source. Landfill gas was an early success of British renewables policy, and one of the few forms of renewable energy in which Britain genuinely led the world. Containment and collection have therefore been increasingly efficient and complete for many years. The figures for landfill gas electricity production therefore represent an increasingly accurate picture of the gas produced from the putrescible material in the landfilled waste streams.

Gas is produced much more slowly in a landfill than in a digester, of course. The production pattern is thought to be roughly three years of rapid exponential growth to peak production, followed by slow exponential decline for decades. But if one adds the output from cells in the first year of their production to those in their second year to those in their third year, etc., the annual output is similar to the output if the material were digested rapidly in an AD plant, once enough years of waste are within properly-engineered cells.

One has to make allowances for some differences. The academic and regulatory opinion on the proportion of gas that has been captured by landfill-gas extraction has declined over the years, to the significant benefit of the government’s calculations of its performance in reducing greenhouse-gas emissions. This was largely a change in the assumptions, as the quantities of fugitive gas are not amenable to measurement.

But landfill gas has a distinctive odour and most people know when even small amounts of it are present. Gas migration is therefore usually detected and dealt with promptly because of the regulatory threats from inaction. Usually, if you smell landfill gas in the air, it is obviously coming from the open face of the landfill, not seeping out of the ground some distance from the landfill via a leak in the liner.

A more reasonable assumption than is applied nowadays is that most of the gas is captured once a cell is sealed. Extraction is required and incentivised, creating negative pressure within the fill, so gas is unlikely to leak in significant volume even if the liner is damaged. We may assume that a significant proportion of the gas from an open cell escapes. But gas production in the first year or two is a small proportion of the total gas that will be released, and cells will typically be finished in that timescale.

AD plants convert the feedstock to gas more efficiently than landfills, so one may allow for a greater capture of gas by AD because of more complete conversion as well as more complete capture. On the other hand, the largest proportion of putrescible material in a landfill is slowly degrading (e.g. paper, card and wood), and it is likely that this will contribute to some extent over (say) 30 years of landfill-gas extraction, but won’t contribute to AD because it will not normally be used as a feedstock.

Not all of the potential feedstocks for AD were going to landfill previously, but some of the key ones were. We can exclude manure, slurries and other agricultural waste (mostly), and of course sewage. But most of the food and biodegradable waste was going to landfill. Some municipal food waste may now be collected separately and sent to AD, which was previously burnt as part of the mixed municipal waste stream going to Energy-from-Waste plants. But the rise of Energy-from-Waste is a recent phenomenon in the UK, and the majority of food waste is not municipal and was never going to EfW.[1] Larger quantities of municipal biodegradable (non-food) waste may have gone to EfW, but as discussed above, this was always more suitable for thermal than biological treatment if we have inferred its definition correctly.

Image

The uncertainties are not orders of magnitude. We may allow for AD producing around 25% more gas than from the same material being landfilled, give or take another 25% (i.e. 0-50%). But the uncertainty over the conversion and capture efficiency is whether it is 60 or 80% not 10 or 40%, given the processes and incentives involved. And the uncertainty over differences in feedstock is also marginal not exponential. We will not be an order of magnitude out if we estimate the potential of municipal and commercial food waste and any other putrescible components of biodegradable waste on the basis of the landfill-gas figures at the peak of national production, before the efforts to divert putrescibles from 

Image

landfills really took hold.

Landfill gas production peaked in 2011 at 1,744 ktoe, but if 2007 or 2008 were the last figures available to E&Y, it wasn’t too far off its peak, at 1,540 ktoe.[2]

It could reasonably have been predicted to be peaking. The Landfill Directive was introduced in 1999, with an objective to divert reusable material (including putrescibles) from landfills. As the previous chart illustrates, volumes going to landfill started to decline from around 2003, although initially that was more about the diversion of conventional recyclables rather than energy recovery. But increased separate collection of food waste and energy-recovery through AD or EfW was on its way in order to comply with the Directive. It was a trivial assessment to judge that gas production was close to its peak (given the roughly 3-year lag between filling and peak production). For example, Summerleaze sold its landfill-gas generation business in early 2007, partly because a decline in production was expected in the not-too-distant future.

One could therefore do a simple calculation to estimate the gas that would be produced if the putrescibles being landfilled were digested instead. If one assumed that the latest figures available in 2009 represented the peak, and adjusted upwards by 25% for greater efficiency in AD, the annual potential of these types of feedstocks was around:

1,925 ktoe (22.4 TWh), within a range of 1,540 ktoe (17.9 TWh) to 2,310 ktoe (26.9 TWh).

By now, we can see that it might have been a little higher:

2,180 ktoe (25.4 TWh), within a proportionately similar range (20 – 30 TWh).

To this one would need to add the biomethane potential of sewage gas and farm slurries, and the potential for gasifying solid feedstocks. As explained above, the realistic potential of each of these for material quantities of biomethane was very limited. History has substantiated what should have been a reasonable expectation in 2009. None of these technologies is contributing material quantities of biogas to our gas networks.

We can discount sewage gas because the economic opportunities are mostly deployed to the most practical use: CHP for on-site requirements.

The problem with manure and slurries is and always was the economics of a low-gassing feedstock with a cost rather than a gate fee to import. As the Ricardo report cited above notes, it is most likely to be used as one of the minority feedstocks, supplementing the main gassing potential of food and other putrescibles in the commercial and municipal waste streams. If the amount of manure increased the total feedstock (by mass) by 50%, it might add 20% to the amount of gas to be expected from the main feedstocks.

Gasification for grid injection is about the economic maturity and cost of the technology. As discussed above, any reasonable assessment would have been extremely cautious about the prospects of this technology. Gasification itself had failed to develop a commercial offering after decades of research and high expectations. In the case of grid injection, we must also consider the immaturity and cost of the technology to convert gasification gas (CH₄, CO and H₂ plus non-combustibles) to methane, or the establishment of a standard and infrastructure in every building to revert to a modern version of town gas.

Though we are as sceptical about the claims for hydrogen as biomethane, we may concede that, if one were going down this route, it would make more sense to convert the gases and network to hydrogen, as it facilitates CCS, rather than go to all this expense to produce a “green” gas whose carbon could not be captured. Hydrogen was already perceived in 2009 as a likely long-term option for heat, as illustrated by the quotes above. There was little reason for E&Y to include gasification gas within the potential for biomethane, other than that they needed it to make the total sound significant enough to affect policy.

So we get a range for the biomethane that could be produced if all the feedstocks in the report that were credible and economic were used for that purpose of around 25 – 36 TWh, i.e. around 2.3 – 3.3 bn m³ of biomethane p.a. As E&Y anticipated gas demand in 2020, this represents around 6.6 – 9.4% of domestic gas, or 2.4 – 3.4% of total national gas demand.[3]

These are effectively the realistic versions of the “stretch” scenario, i.e. using all the feedstock that is realistically available and economic. The “baseline” scenario assumed a material amount of the feedstock continued to go to other uses. For instance, the figures above, like E&Y’s “stretch” figures, assume that no putrescible material remains available for power and/or heat production from AD (other than sewage), and all the current output of that technology would be lost.

Alternatively, in the “baseline” scenario, we may assume that much of this feedstock continues to be used for the existing purposes. We therefore retain much of the 2.7 TWh of electricity generated by AD, but can only rely on a minority of the biomethane calculated above, perhaps 10 TWh (around 1bn m³ or 1% of national gas demand).

Would the government have worked hard to deliver a technology that was likely to deliver <1% of our gas requirements, and might stretch to 3.5% at best? What was the long-term decarbonisation strategy that justified heavy subsidy to deliver this marginal contribution?

This figure could be increased materially if one assumed a substantial contribution from energy crops for digestion, not considered as an option by NG/E&Y. It represents the discrepancy between the 3.3bn m³ upper limit by this method, and the 6bn m³ suggested in the analysis of the individual feedstocks above. The proportions nevertheless remain too small to make a persuasive case that the gas grid could be decarbonised to a material extent.