Thinking is uncomfortable

Most human beings, most of the time, find it much more comfortable and satisfying to react rather than to think.

This should not be surprising: when thinking thoughts spontaneously appear in your mind, and until you have acquired the mental tools to deal with them you don’t know how to manage them: they seem to be pulling in lots of different directions, meaning many things, and without the habits to arrange and manage them, it is all just too confusing. Reacting (the ‘gut reaction’) is just so much simpler: it gets straight to an answer without any of that confusion.

So why do humans think at all? And why do some of us seem to be stuck in not being able to stop thinking? This has the familiar aspect of a biological signalling mechanism which has got ‘stuck’ in the ‘on’ position, by some mutation, when it had evolved to only fire in response to a particular set of stimuli. But I don’t think this is enough to distinguish Homo sapiens sapiens from our progenitors, since most of us today are not ‘stuck’ and very rarely indulge in ratiocination at all (a favourite term of Peter Medawar‘s).

The mental tools to make sense of our own thoughts are very basic: a toddler will learn that when something pushes from the left, it will fall over unless also pushed from the right: thesis and antithesis. And useful lists have been compiled (by Aristotle and many others) of when these tools go wrong, and the world might be a better place if primary schools covered all of these before a child is 10.

With these tools in place, more of us may be more comfortable thinking than reacting, and that would be a good thing. Philosophy in schools seems to often be more concerned with moral philosophy and difficult things like that, and not be covering the basic tools that are essential for a thinking mind.

Oxygen boilers – making condensing boilers more efficient

A condensing boiler is more efficient because it recovers some of the latent heat from the hot water vapour in the flue gas. However, no domestic condensing boiler manages to condense all the water, and how much is condensed depends on the setting of the temperature of the condenser, which is typically the ‘return temperature’ of the water in the central heating system.

Fig.1 Condensing boiler maximum possible efficiency for two different fuels, natural gas and hydrogen, as a function of the condensing temperature. Fuel gas is at 8C and the inlet air temperature is at 5C. There is excess air of 15%.

Fig.1 shows how the efficiency varies with condensation temperature. (Hydrogen is slightly better at low condensing temperature, and quite a bit worse at higher temperatures.) Note that for natural gas there is no condensation at all above 55C. This is controlled by the partial pressure of water vapour which is entirely determined by the chemistry of the fuel and the quantity of inert non-condensible components in the flue gas: nitrogen, carbon dioxide and a trace of argon.
One might hope that by adding a little hydrogen to the natural gas, one might shift the dew point (the temperature at which water condenses out of the flue gas composition) to get more efficiency. Unfortunately this is a very small effect indeed, as can be seen in Fig.2

Fig.2 As Fig.1, but with an additional curve for a fuel which is the same natural gas, but with hydrogen added to make 20% of the total (by volume).

Clearly, to make a big difference we need to do something about the ~80% inert gas in the inlet air, as it is this which is diluting the water vapour and preventing it all condensing. Fig.3 shows that this is a significant effect, but condensation is still limited by all the CO₂ which is produced by burning the natural gas.

Fig.3, natural gas maximum condensing boiler efficiency as in Fig.1 but with varying amounts of oxygen in the inlet air. Normal air has 20.9% oxygen.

There are practical limits to how much nitrogen should be removed. As progressively more nitrogen is removed, the efficiency goes up but also the adiabatic flame temperature increases. This increases the amount of poisonous nitrogen oxides produced. At 100% oxygen, this is not a problem because there is no nitrogen to be oxidised: but producing very pure oxygen is likely to be much more expensive, and the flame temperature would be so high that burners would need to be made from very expensive materials.

Oxygen-fired combustion is well understood in large industrial furnaces and the usual solution is to recirculate some of the flue gas into the flame to reduce the temperature. This may be feasible for a domestic boiler, but would increase the cost of manufacture.

We can get rid of the carbon dioxide by using a fuel which only produces water vapour: hydrogen, but we would still have the nitrogen oxides problem without flue gas (water vapour) recirculation. The efficiency gain appears to be quite significant, see Fig.4.

Fig.4 Maximum boiler efficiency for a condensing boiler burning hydrogen.

Fig.4 shows the very substantial increase in efficiency that occurs with oxygen enrichment of a hydrogen condensing boiler. This is such a large effect that one might expect all such boilers to employ some degree of oxygen enrichment, as (a) this would reduce the usage of the (quite expensive) hydrogen gas, and (b) hydrogen boilers need to be completely redesigned anyway, because of the different flame velocity.

Note in Fig.4 the small segment of sensible heat loss even for 100% oxygen near 100C. This is due to the excess oxygen which is not consumed by the fuel: it carries off some of the water vapour which would otherwise be condensed. All these calculations assume 15% excess air (or air mixture) to ensure complete combustion. (With a more sophisticated control system, this excess could be reduced after the boiler has started up and is running continuously, as is standard on large industrial systems.)

Figs. 5 and 6 show the same data, but as difference from the case where the fuel is burned in normal air. These figures show two kinks: one at the dew point of the reference case, the fuel burned in air, and the other at the dew point of the fuel+air mixture. These are not proportional percentage increases, these are absolute increments: the efficiency of burning natural gas goes from ~91% in air to ~94% efficient in 30% oxygen at 50C, an increase of ~3% (blue lowest line in Fig.5).

Fig.5 The data in Fig.3 replotted as difference from the reference case of burning the fuel in normal air.

Fig.6 The data in Fig.4 replotted as difference from the reference case of burning the fuel in normal air.

For more detail, and to see the text of the patent and links to other reports, see https://oxyboilers.co.uk .

[ Update 31 March 2024 ]

Alternatively, increase the pressure

If the condenser is at a higher pressure than atmospheric, then the partial pressure of the water vapour is too, so more of it will condense. Which makes sense intuitively as it “wants” to be in a higher density state when the pressure is higher. It doesn’t take much to shift the dew point quite usefully:

Alternative, ideal flameless boiler

The ideal solution to recovering all the energy in the water vapour would be to use a flameless cell which burned the fuel on a catalyst at the temperature of the central heating return water, or an electrochemical cell which did the same thing. The water would never be in the vapour phase, and so would not need condensing.

The code that calculates these figures is all public on GitHub. It is written in python.
The extension of this treatment to LPG boilers is left as an exercise for the student.

Hydrogen in pipes

A paper about to be published completely revises previously held understanding on how hydrogen flows in low pressure pipes, such as the final few hundred metres of the domestic gas distribution network ( which would be relevant if we ever replace any natural gas with hydrogen).

Since 2004 we have thought that we understood this. Now is is apparent that we did not.

Does this affect plans for using hydrogen domestically ?

Not really. The final numbers don’t change that much. It is only at periods of quite low gas demand that the pumping power required to deliver calorific value via hydrogen is nearly 8 times what it was for natural gas:

Fig.1 For five different pipe roughnesses, the ratio of the pumping power required to make up for friciton losses for hydrogen compared with natural gas. The Reynolds’ number is a measure of gas velocity. Most service line pipes have Re < 3000. [Actual compressor power required will be different depending on compressor efficiences.]

What is more worrying is that this work makes it clear that everyone working in this field for the past two decades has made assumptions about gas flow which are quite incorrect. And the concern is that modern process systems computer codes, which are perfectly fine for modelling process plant, may be using the wrong calculations for slow-velocity, low pressure pipes such as the domestic distribution grid. These codes are black boxes and users have no way of finding out how the calculations are done.

Everyone has assumed that Reynolds’ numbers more than about 2,000 mean that the flow is turbulent. This is not right: for a smooth pipe such as polyethylene used today (relative roughness ϵ/D ~ 10^-5), the flow in a circular pipe is only fully turbulent above a Reynolds’ number of more than 10⁹ . It stops being laminar at about 2,300 but then it is in a Blasius and pseudo-Blasius smooth-turbulent regime (shown in the figure by the purple line) for 5 orders of magnitude of gas flow velocity.

There have been very significant advances in the understanding of viscosity in the past 30 years and previously people working in hydrogen techno-economics seemed to have been using old text books, not current research knowledge.

The final numbers do change slightly from what had previously been estimated (estimated for pure methane incidentally, not for natural gas which is a bit different). Hydrogen needs a much higher gas velocity in the pipe to deliver the same useful energy, about 3x, but the exact number we now know is 3.076 times faster (for flows where both the natural gas and hydrogen are in the Blasius regime, something that is actually quite unlikely and rare in the final connection to domestic housing).

Because hydrogen has a low viscosity, the extra gas velocity does not need as high a pressure drop to drive the velocity increase as one might expect. The correct number is that a pressure differential 1.290 times higher is required (in the Blasius regime). This is a bit of a stretch, while the vast majority of distribution pipes will be fine (these run between 75 mbar and 20 mbar above atmospheric pressure), a very few already near the limits of their capability to deliver natural gas will not be able to deliver enough hydrogen.

Unfortunately, at the low gas velocities in the final metres of the distribution network this number can rise to 2.5 times higher, but this is only at times of low demand when there should be adequate pressure gradient ‘headroom’. However it is worrying that gas engineers seem to be totally unaware of this issue, or at least, nothing public has been published about it.

The pressure drop multiplier, as a function of gas velocity, has the same form as Fig.1 above:

Fig.2 How much higher the pressure gradient has to be to delivery the energy-equivalent amount of hydrogen.

So why has the gas delivery industry been so out of date? The Moody diagram which describes the flow of fluids in pipes dates from 1944, but the oil and gas industry has historically used a wonderfully bizarre menagerie of oil-industry specific codes and formulae, such as the Pandhandle equations¹ developed for gases in Texas: these are of course completely inappropriate for hydrogen as they have been calibrated quite specifically for certain natural gas mixtures only.

Fig.3 Updated Moody diagram, showing the laminar/turbulent transition, but also the ‘belly’ and ‘rise’ of the ‘Nikuradse behaviour’. Nikuradse was Prantl’s research student: this is originally from the 1930s.

This has all been re-measured and re-calibrated over the past two decades with two machines at Princeton and Okinawa. There has also been significant theoretical development by Goldenfeld at the University of Illinois.

Fig. 4 The maximum possible boiler efficiency for an air temperature of 5°C and a gas temperature of 8°C for varying condensation temperatures (similar to the return temperature on a central heating system)

The paper also incidentally calculates the form of the boiler efficiency curve for natural gas and for hydrogen, which is perhaps much more significant for those planning on using hydrogen for domestic heating.

Fig.4 shows that a hydrogen fuelled boiler is much less efficient, by about 5% above 70°C, than a natural gas boiler if the condensation temperature in the condensing boiler is warmer than about 63°C. However as the central heating return temperature is decreased, a hydrogen boiler will be slightly more efficient that a natural gas boiler at the same condensation temperature. (This extra efficiency at 50°C was taken into account in the paper when calculating the required hydrogen velocity.)

So a hydrogen boiler really does need to be coupled with a weather-compensation control system in every installation to ensure that hydrogen is not wasted, and it shows that using hydrogen is not a way to avoid having a strong incentive to reduce the flow temperature of your radiators.

Fig.5 The gradients of the lines in Fig.4 showing the strong peaks of efficiency improvement when the return temperature drops just below the dew point temperature for each fuel.

Fig.6 The ratio of friction factors for hydrogen and natural gas to deliver the same useful energy: an enlargement of the flow regimes for most of the gas distribution network. This is not quite the same as the ratio of pressure drops, because of the difference in viscosities and densities.

[minor edits 2024-0320]
Okinawa device size and axes labels on ‘friction loss’ plot, which is now at 30 bar.
Also scfd is a mass unit, not a volume unit.

1. “Where units are in scfd, degrees Rankine, psia, feet, and miles“. scfd is ‘standard cubic foot’ per day, a measure of gas mass delivery. Degrees Rankine have the same interval as degrees Fahrenheit, but with zero at absolute zero: °R = °F + 459.67. psia is pounds (force) per square inch absolute, i.e with respect to zero pressure not with respect to atmospheric pressure. No wonder they got confused.
   ↩︎

Hydrogen embrittlement for economists

Over the past decade there have been an overwhelming number of techno-economic studies of hydrogen deployment as a means to reduce national carbon emissions. To reduce cost, these have always considered the possibility (or necessity) of re-using existing natural gas pipe networks.

Far too many of these economist-written reports contain an (unreferenced) of-the-cuff warning about hydrogen embrittlement, and these well-meant (but ignorant) warnings are then picked up by campaigners and lobbyists with no respect for truth or any evidence whatsoever, only rhetorical scoring points.

Some years ago I was so appalled (I am a metallurgist and was a Fellow of the UK’s ‘learned body’ the Institute of Materials, Minerals & Mining and also a chartered materials engineer) that I tried to find out the source of all this misinformation. I discovered one them: the Wikipedia entry on hydrogen embrittlement had been so severely polluted by hearsay, modern myths and unsubstantiated gossip that it needed serious attention. I and several other Wikipedia editors worked on it for a while and today it is in much better shape (though the reading-age required might still be beyond many journalists I fear: more work is required). But it is interesting to note that every few months some poor well-meaning soul edits it to add more biased and un-evidenced garbage, so keeping this clean requires constant effort.

This is particularly galling since

• the basics of this embrittlement have been well understood since 1875
• ordinary steel is today, and always has been, the material almost universally used for compressed hydrogen cylinders (at 300 bar!)
• much of the existing UK iron pipe network has already carried coal gas (50% hydrogen) for many decades

This is not to say that embrittlement can be ignored. New networks proposed using higher-strength steel alloys certainly need to use fracture mechanics for correct design, and rotating machinery in hydrogen atmospheres needs to be specified while understanding the environmental effects on crack growth rates. The danger from hydrogen embrittlement in pipelines comes from welding the pipe with damp welding rods, from acid cleaning, or even from external rusting, but not from the hydrogen inside.

In the past 2 or 3 years the level of stupidity of this sort has seemed to reduce, so it looks as if correcting Wikipedia had a useful effect. There were still some articles published, but mostly from local newspaper journalists, or geographers or economists (who couldn’t be expected to know any better), and hardly any from science journalists (who certainly could be expected to know better, or at least to check with Wikipedia) and even a gratifying reduction in the number from outraged loons (who wouldn’t care about knowing better).

Until last October.

Physics World published an otherwise-excellent article on heat pumps which also discussed hydrogen for heating and unfortunately contained this phrase “… would require not only new boilers but also replacement pipes, since hydrogen in high concentrations causes steel to become brittle“. This is not just condescending side-insult, a wink and a nod, or a sly insinuation of trouble if you go in that direction, but an outright and complete untruth. What is more, neither the writer nor the journal editor thought it important enough to quote or to find any supporting evidence whatsoever. Such a level of ignorance and self-importance in thinking that this is unimportant is staggering. If this were the parish journal of the Outer Boondocks Chronicle and Advertiser, I would just sigh and move on, but this is the institution journal of one of the most respected learned societies in the UK. This article will no doubt be quoted as authoritative evidence by anti-technology campaigners for decades to come: the damage has already been done.

There are two authoritative sources I recommend, in addition to the Wikipedia article:

• Prevention of Hydrogen Embrittlement in Steels, by Harry Bhadhesia (2016). The first two pages of this long review article.
• Technical Reference for Hydrogen Compatibility of Materials, Sandia Report SAND2012-7321 (2012), which has all the numbers you could ever need. Though you may need a masters degree in linear elastic fracture mechanics to be able to use them properly, or at least understanding why the units for fracture toughness have square-root(metres) in them: MPa m^1/2

[ Addendum: 31 March 2024 ]
ASME standard B31.12-2014 unfortunately contains these words: “Hydrogen gas molecules adsorb onto metal surfaces, like many atmospheric gases, and dissociate to its atomic form.” This is entirely and completely incorrect in every particular. The dissociation energy of hydrogen is huge, 423 kJ/mol, so this Simply Does Not Happen (not unless the pipe is glowing red-hot, in which case other failure mechanisms may be more worrying). Check with Prof. Bhadhesia’s review article.
This ASME standard, quite rightly (from a position of ignorance) mandates a large safety factor to be used with things that the revision team did not understand. This does not mean that the “materials performance factors” they mandate have any connection to reality. I look forward to the next revision of this standard.

[ Addendum: 3 March 2024 ]

While a home electrician installing a solar hot water system in Outer Boondocks, Lower Nebraska, may honestly believe that the broken weld he is looking at is due to this mysterious ‘hydrogen embrittlement’ thing, and while his distant cousin in Blackstumpton, Eastern Australia (who is a bit hazy on the distinction between atoms and molecules) may encourage him in this belief¹, this is not objective evidence backed up by careful, long-term experiments, deep thought by a community of clever people, and a lot of electron microscopy.
That is the the level of the material which was excised from Wikipedia.

This is what Wikipedia says now: “Gaseous hydrogen is molecular hydrogen and does not cause embrittlement though it can cause hot hydrogen attack (…). It is the atomic hydrogen from chemical attack which causes embrittlement because the atomic hydrogen dissolves quickly into the metal at room temperature.”

It is worth noting that the fast fracture²of a specimen filled with atomic hydrogen is an effect with origins in the bulk of the specimen material, whereas the effect of gases on fatigue crack growth are (almost certainly) due to the adsorption of gas molecules on the freshly-formed metallic surface at the crack tip on each cycle of crack growth. The term “hydrogen embrittlement” refers specifically and only to the bulk effect.

1. I do not wish to cast aspersions on the many fine universities which no doubt exist in Eastern Australia, but Bruce (for that is surely his name) has never been to any of them. ↩︎
2. “Fast-fracture” is a technical term. It is quite different from fatigue crack growth. Look it up. ↩︎

Carbon Dioxide Removal – CDR

Greenpeace recently published an excellent briefing on several options for directly removing carbon dioxide from the atmosphere. There are a couple of small, but notable mistakes, and the way they have used upper- and lower-bounds in the main result diagram needs a better approach. One of the quibbles changes the whole meaning of part of their report.

Here are the main readiness & value conclusions of Greenpeace’s study :

Copy of the figure from the Greenpeace report

Note that they conclude that inorganic CDR works. There are no engineering show-stoppers, but it is quite energy-expensive. It requires a workable but high carbon price. Governance of inorganic methods is also much easier as tonnes of liquid CO2 (DACCS) are very measurable. Ocean liming is probably much cheaper (though they don’t mention that) but it and the weathering and fertilisation methods are more indirect in measuring the CO2 stored. These conclusions are unfortunately not reflected in the colour diagram above.

A major conclusion in the text is that the governance and baseline issues that we have seen with REDD+ and CDRs in the Kyoto mechanism mean that organic methods of generating carbon credits from forests, soils and biochar are very unlikely to work in practice. This important finding is not reflected in the colour diagram above or in the summary and conclusions of the report.

Additionality

Although the report mostly separates the different CDR technologies admirably, some sentences lump several together in a misleading way. So where the notion of additionality is discussed (page 10), it lumps Direct Air Capture (DAC) in with the others without saying that most additionality problems are really not an issue for DAC whereas it is for the others. Though the potential issue of possible double-counting of DAC in with national targets still remains.

Framing

There is an implicit framing to the whole article which means that parts of it are misleading. The implicit framing is that we are comparing technologies over the whole period between now and 2050: it addresses the degree to which each tech can make a difference over that period. They key framing is “over that period”. Thus without explicitly saying so, the reader is lead to compare technologies in terms of total megatonnes and gigatonnes of CO2, not in terms of annual rates (which will be varying dramatically over that period).

Optimism and pessimism: upper and lower bounds

When doing work of this sort there is useful prior experience on where to be pessimistic and where to take upper-bounds when summarising technology attributes. The UK civil service has a great deal of practical experience with using ‘red’, ‘amber’, ‘green’ RAG codes in assessing attributes of ongoing projects, sometimes including a ‘black’ category for a project which currently cannot possibly achieve its goals (!), and we should use that experience.

These sorts of reports are most useful in prioritising today’s resources. So they need to be ultra-realistic, i.e. pessimistic, taking a lower-bound, about whether a technology can be deployed immediately. This is because immediate large-scale deployment is very expensive. This is the first column of the diagram labelled ”Technological readiness’.

But they should be open to all possibilities when looking at long-term potential ‘CO2 potential’ because the only resources being immediately deployed would be research – which is cheap. So it is most useful if to take an upper bound, i.e. be optimistic, when talking about long-term potential.

So while I would argue that the ‘central estimate’ for DACCS should be amber or green and not red (since a plant in Switzerland is already operating), I am somewhat relaxed that the authors of the report are more conservative and have classed it as ‘red’. However I disagree most strongly about the classification of Soil Capture as ‘green’ because while I agree that the central estimate is somewhere in the green-amber spectrum, for methodological reasons one should be conservative and therefore class it as ‘amber’.

On this logic one can see that the other columns ‘Cost’, ‘Permanence risk’ and ‘Environmental impacts’ should all be pessimistic or optimistic depending on how they affect short- or long-term resource use. But it is not so simple: they all have mixed influences so one can’t make a simple judgement. Arguably ‘Cost’ should be a long-term, thus optimistic measure. When one has attributes in a diagram that are mixed-up in this way it is an indication that one should look for better attributes which have a clearer interpretation. Renaming ‘Cost’ as ‘Expected 2050 cost’ would improve clarity for example.

Finally, there is a well-established scale for technological readiness levels: TRL between 1 and 9. But it is only really applicable to very specific innovations, not to a whole class of similar technologies. Nevertheless it would be useful to at least attempt a TRL coding on this diagram. This sort of thing is actively used when assessing projects today, e.g. for the Energy Entrepreneurs Fund grants .

Major quibble

The ‘CO2 potential’ is the total amount globally that could realistically be removed from the atmosphere by each method. It is explicitly wrong in the colour diagram for the last three methods. Greenpeace have misquoted their reference to the IPCC (SR15) and Fuss et al. (2017) which clearly state that there are effectively no technical limits on the amount of CO2 that can be removed by rock weathering, ocean fertilisation and ocean alkalinisation: those three blocks in the second column at the bottom of the diagram should all be green. The Greenpeace report has applied a comment in the source documents which was only describing some technologies as being relevant to all of them. IPCC SR15 Box 7, Table 1 (p270) doesn’t mention the two ocean methods, and the comment there about weathering only applies to rock dispersed on soils. So there is no cited evidence to support the three coloured rectangles for those technologies:

No evidence for the two bottom parts of the diagram, partial evidence for one type of weathering only for the top one.

The rest of the text in that section “Maximum sustainable potential” applies mostly to the land-using AR and BECCS technologies but is written as if it applies to the bottom three. It doesn’t. (The Fuss review paper is where the IPCC Sr15 gets most of its data from anyway so these are not really separate sources of data.)

Note that the report states an upper limit of 5 GT/y for all DACs (quoting the Fuss report https://iopscience.iop.org/article/10.1088/1748-9326/aabf9f/pdf… ) but that is a mistake. Fuss gives no upper bound for ocean liming at all. The limit on tech potentials are for other techs. Fuss et al. : “if DACCS becomes competitive, potential deployment will be driven by cost support and rates of upscaling, with no obvious upper biophysical limit, barring storage, material and thermodynamic constraints” (Negative emissions—Part 2: Costs, potentials and side effects). That implies that those three rectangles should be pure green.

The purpose of IAMs and why they seem to like BECCS

The point could be made that many, probably most, IA Models use BECCS as a “placeholder” technology. It represents a bunch of different ways of removing CO2 from the atmosphere at a particular cost. That is all the level of depth that the model needs to know (most do not represent land us explicitly) for its purpose.

So when a IAM publishes results that say it requires a lot of BECCS, what it is are really saying is that their scenario requires a lot of ‘something that directly removes CO2 at the same price that we have guessed for BECCS’, not specifically BECCS itself at all. Certainly UK-TIMES does that (from personal experience), but the results are published by people who understand that – so there is no confusion in practice. IAMs however are re-used by people who have no direct experience of how the model was constructed.

Conclusion

This is a very worthwhile report. I just wish it could have been a page or so longer so that those quibbles and framing issues could have been made more explicit.

Very recent articles on CDR, GGR, CCS, NETs

Everyone working in climate change should at least skim-read the CDR Primer.

HuffPost article “Should We Spend $1 Trillion A Year On This Climate Technology? We May Have To.” was updated 14 Jan. 2021.

Article in The Engineer: “Direct air capture: silver bullet or red herring?” 13 Jan. 2021. Ciaran McKeon, of Frazer-Nash Consultancy, explores the environmental potential of direct air capture technologies “DAC systems, are less limited by location, require less space and use less water than other negative emissions technologies”.

Iliana et al., (2013): “On short time scales, alkalinity enhancement could target local ocean acidification mitigation allowing avoidance of large changes in surface pH and Ω in certain high‐value ecosystems of the world ocean” in Assessing the potential of calcium‐based artificial ocean alkalinization to mitigate rising atmospheric CO₂ and ocean acidification.

Footnote

I could strangle whoever reformatted the Greenpeace report though: they pushed all the excellent, thoughtful footnotes to the end, thus ruining the in-context commentary depth. For example note [125] on page 13 is not a citation to past work, but a direct explanation of why IAMs favour delayed action. They also changed the citation style from (Author, year) to [n] thus wrecking the intra-citation references because they did not also change the citation format within the cross citations to match.

Extra references (added 22 March 2021)

Assessing ocean alkalinity for carbon sequestration (2017) https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016RG000533

BEIS Greenhouse Gas Removals – Call for Evidence (2021)
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/941191/greenhouse-gas-removals-call-for-evidence.pdf

Antacids for the Sea: Artificial Ocean Alkalinization (2020)
https://legal-planet.org/2020/01/27/antacids-for-the-sea-artificial-ocean-alkalinization/

Biomethane upgrading plant and fugitive emissions of methane

Tomorrow, July 7th 2020, is the deadline for commenting on two government consultations affecting biomethane subsidies:

• “Non-Domestic Renewable Heat Incentive – Ensuring a sustainable scheme”, which is about the extension of the non-domestic RHI for a year which is where the current subsidy for biomethane comes from,
  https://beisgovuk.citizenspace.com/heat/ndrhi-2021reforms/ and
• “Future Support for Low Carbon Heat“, half of which is a proposal for an entirely new policy to take over the financial support for biomethane in the UK.
  https://beisgovuk.citizenspace.com/heat/future-support-low-carbon-heat/

Both of these include a “value for money” calculation (VfD) using the Treasury guidelines to calculate the overall benefit of the proposed policy to the country as a whole. This VfM allocates costs for emitted greenhouse gases based on the current government cost for carbon. Since BEIS is a government department they don’t have any choice about that. But they do have control over what they include and exclude from that calculation.

There are three significant omissions which mean that the calculated VfM and thus the benefit of the policy as a whole are likely – not just possibly – but likely to be significantly in error:

1. methane emissions from the AD plant by design
2. fugitive and accidental emissions of methane
3. using English historical data to predict counterfactual future food waste disposal emissions when there are regulations in place that will make this very different in years to come

Fixing these is relatively cheap and should be an essential part of future government policy. A handful of plants already capture all the CO2. These numbers show that all of them should. The very best plants might then even be slightly carbon-negative.

It hardly needs saying that methane emissions over about 2% also mean that the climate benefit of the technology as used in the UK is severely in doubt. We may be spending public money to make the climate worse; so we need to do these calculations very carefully and with full public scrutiny. This problem is at a level of detail that we cannot expect parliamentary scrutiny to be much use.

1. Emissions by Design

The first omission is the intentional and designed-in release of methane to the atmosphere by the biomethane separation “upgrading” plant. This takes the biogas from the digester and separates it into biomethane and CO2. The CO2 is vented to the atmosphere. But no separation plant is perfect so some methane is always emitted too. This is a non-trivial and very climate-significant amount.

All the data to know what these emissions will be should be in the plant design specifications when the project is initially proposed for support.

“Methane slip from biomethane upgrade plants is typically around 1%-2% of the total methane production, although this can be up to 10% and as low as < 0.1% . Technology choice and economics are key drivers for emissions from biogas upgrading with technologies like cryogenic separation requiring a large amount of energy. The cost of the technology can influence the efficiency of recovery”

“Chapter 9 – GHG Emissions From Biomethane Gas-to-Grid Injection via Anaerobic Digestion“, Adams, University of Bath, (2017) https://www.sciencedirect.com/science/article/pii/B9780081010365000094

This University of Vienna/ Intelligent Energy report “Biogas to biomethane technology review” (2012) is an easy to read review. Two of the commonest separation technologies in the UK cannot emit less than 2% of the methane (page 13) and require a catalytic converter to fix this. Such catalytic converters are not mandated in UK regulations.http://www.severnwye.org.uk/fileadmin/Resources/SevernWye/Projects/Biomethane_Regions/Downloads/BiogasUpgradingTechnologyReview_ENGLISH.pdf

This is one of many such studies as it is extremely well-known, so much so that its omission from the BEIS consultations is at best incompetent, e.g. “Methane emissions from biogas plants”, IEA Bioenergy Task 37 (2017) 12 https://www.ieabioenergy.com/wp-content/uploads/2018/01/Methane-Emission_web_end_small.pdf

2. Fugitive and accidental methane emissions

Fugitive and accidental emissions are a big concern because of poorly-trained farm staff not understanding the climate implications of leaks.

Fugitive non-accidental emissions are mostly from the operation of pressure release valves (PRV): “Large quantities of methane emissions have been reported caused by single large leaks or long lasting pressure relief events. … The application of specific monitoring and maintenance and/or the application of specific technologies can reduce emissions. A crucial part of any operation should be a monitoring plan .”

(a) http://task37.ieabioenergy.com/files/daten-redaktion/download/Technical%20Brochures/IEABioenergy_Task%2037_methane_emissions_2Psummary.pdf

The PRV methane emission, which is the plant operating as designed,  has been measured at 3.88% in one plant – this is in addition to all the other emissions and accidental and maintenance releases. Analysis of operational methane emissions from pressure relief valves from biogas : “the determined methane emission factors are 0.12 g CH4 kWhe−1 (0.06%
CH4-loss, within 106 days, 161 triggering events, winter season) from biogas plant A and 6.80/7.44 g CH4 kWhe−1 (3.60/3.88% CH4-loss, within 66 days, 452
triggering events, summer season) from biogas plant B”. https://www.sciencedirect.com/science/article/abs/pii/S0960852416302103
The average emission of all the plants studied in that paper was 4.6%.

“Methane emission rates varied between 2.3 and 33.5 kg/hr CH4, and losses expressed in percentages of production varied between 0.4 and 14.9%. The
average emission rate was 10.4 kg/hr CH4 and the average loss was 4.6%.
https://www.sciencedirect.com/science/article/pii/S0956053X19304842

An Irish report (2020) assumes AD plant emissions with a standard 2.4% CH4 fugitive loss allowance:
https://www.sciencedirect.com/science/article/pii/S0048969720330114?via%3Dihub

The climate-cost of a methane leak is out of all proportion to the lost revenue by the AD plant operator, so this matter has to be managed by regulation, minimum standards, inspection and monitoring rather than relying on the self-interest of the operator to save gas which will be sold – this is well known:

(b) IEA Bioenergy: https://www.ieabioenergy.com/wp-content/uploads/2018/01/Methane-Emission_web_end_small.pdf

Ricardo report to BEIS (2016):  https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/786756/Methodology_to_Assess_Methane_Leakage_from_AD_Plants_final_report_part1.pdf (I was one of the assessors for the bidders to do this project but I left the civil service before the project completed.)

3. Counterfactual food waste CH4 and NO2 emissions

Using English historical analogies for the counterfactual upstream emissions from food waste is simply wrong (impact analysis (IA), annex F) so the “central estimate” is not central, it has a strong bias. As the Welsh example shows (quoted in the IA and also in (a) the NDRHI extension consultation IA BEIS008(C)-20-CG), separated food waste *at worst* goes into the green collection and municipal composting, not landfill. So the “10% to landfill assumption” after separate food waste collection is a mistake.

The NDRHI uses 20% landfill which is inconsistent with this IA, so one of them has to be wrong and in fact both are.

See also the (b) later correction to (c) the POST note https://post.parliament.uk/research-briefings/post-pn-0565/ and this blog article (d) which expands on that correction:

(d) https://philipsargent.wordpress.com/2017/11/29/biomethane-degrees-of-decarbonisation/ which shows that official figures on which emissions calculations are based are also wrong in part.

4. Methane slip from on-site CHP gas engines

The emissions from the  almost-invariably-present CHP biogas engine come under the UK regulations for CHP, but the cost of their emissions is directly applicable to the AD plant and should be attached to the VfM calculation as a monitoring item.

CHP emissions are discussed in http://task37.ieabioenergy.com/files/daten-redaktion/download/Technical%20Brochures/IEABioenergy_Task%2037_methane_emissions_2Psummary.pdf

BEIS is consulting on making minor changes to its CHP regulations (closing date Sept.2020) but from the questions posed it seems intent on not rocking the boat: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/891722/chp-route-to-2050-call-for-evidence.pdf when what is needed is a complete root & branch review and restructuring. The null hypothesis is that zero public money for CHP would give the best VfM policy and has a lot of data in its favour, e.g. this report which they quote but do not seem to understand: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/389070/LCP_Modelling.pdf 

Green gas – Feedstock reporting

Consultation response:

• The consultation and Impact Assessment do not clearly say that energy crops are the most profitable way for farmers (not urban food waste plants) to produce methane. This is misleading as it is important context.
• They do not say that therefore we can expect all plants to use the maximum permitted amount of energy crops.
• The upstream emissions of maize are severe, and while a % rule is administratively simple, it is simply wrong.
• The farmers should be allowed to use whatever maize they want, but pay for the consequence upstream emissions. Defra has a simple calculator for maize emissions (this does not need to be absolutely accurate) which is referred to in the IA, and farmers have to report the % of energy crops already, so it is an extremely simple calculation and easily within Ofgem’s capability to reduce the tariff payments quarterly accordingly.
• If they object to Defra’s standard calculation they can pay a consultant to show that they manage to produce maize without such high emissions – which will give them an incentive to actually do that, which a simple % level does not do.
• So the policy should be redesigned to encourage the behaviour we want in future: to provide incentives for continuous improvement.

Update – 11th March 2021

Bakkaloglu et al. have published analysis of physical measurements of methane emissions from 10 UK biogas plants. The observed losses range from 0.02% to 8.1% of the methane production rate. The average loss was 3.7% which means that burning this gas has the same climate impact as burning coal.  Scheutz and Fredenslund (2019), found an average of 4.6% emissions at 23 biogas plants (in Denmark?).

Update – 23 September 2021

The consultation ended and the government published its response: https://www.gov.uk/government/consultations/future-support-for-low-carbon-heat

They plan to launch a new Green Gas Support Scheme in Great Britain: a tariff-based scheme supporting the injection of biomethane produced via anaerobic digestion into the gas grid. A new Green Gas Levy on
licensed fossil fuel gas suppliers will fund this scheme. Both the scheme and levy are intended to launch in autumn 2021.

As NNFCC say “under the GGSS, producers of biomethane will be required to make a GHG emission saving threshold of 70%.  This sees the greenhouse gas threshold lowered from 34.8 gCO₂ eq per MJ (as it was for the RHI scheme) to 24 gCO₂ eq per MJ. A new methodology will allow for averaging of emissions across consignments and accounts for covers on digestate stores. In addition, feedstocks restrictions that will apply as they do in the RHI – 50% of biomethane must be derived from wastes or residues” which is all well and good, but hardly sufficient.

My concerns about accidental and life-cycle emissions are mentioned and dismissed: “A few respondents suggested including a form of carbon pricing alongside the scheme or instead basing tariffs on a whole lifecycle assessment of carbon emissions and savings from a plant, varying the tariff offered by carbon intensity of the biomethane produced…. However, this represents a significant change to the tariff structure of payments proposed and we do not have evidence available or a standardised assessment methodology to set tariffs in this way”, i.e. a brush-off. There are utterly no measures to detect or penalise methane emissions from operating AD plants whatsoever under the proposed policy.

CIBSE were rather worried about the proposals https://www.cibse.org/getmedia/d5cfa540-9713-4a9b-91ed-fa91433673a0/Consultations-on-low-carbon-heat-and-non-domestic-RHI-CIBSE-response_1.pdf.aspx for reasons entirely unconnected with my concerns about methane emissions.

NNFCC’s thoughts about the eventual policy in full: https://www.nnfcc.co.uk/news-future-support-for-low-carbon-heat

However the Impact Assessment has been revised after the government response, and my points about the food waste counterfactual have been taken into account: “this impact assessment does not quantify carbon savings linked to the diversion of food waste from landfill or other destinations to AD. This has reduced reported lifetime carbon savings of the Green Gas Support Scheme from 21.6 MTCO2e to 8.2 MTCO2e. This conservative assumption is likely to underestimate the carbon savings of the policy” !
It also says “The overall biomethane emissions factor for the policy proposal under our central scenario assumptions is 30gCO2e/kWh which is lower than our previous assumption of -221gCO2e/kWh”. This is in fact an increase in the admitted CO2e emissions per kWh for the future biomethane policy. But it
-omits operational methane emissions completely, which can be up to 10% of delivered biomethane
-omits the possibility of capturing the CO2 and sending it to sequestration

The impact analysis – updated last week 16/09/2021: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1018133/green-gas-impact-assessment.pdf

“

Energy Minister Kwasi Kwarteng confirms 29 [k]g CO2e/MWh applies to future CfD rounds

Minister of State
Department for Business, Energy &
Industrial Strategy
1 Victoria Street
London
SW1H 0ET

Daniel Zeichner MP
House of Commons
London
SW1A 0AA

+44 (0) 20 7215 5000
enquiries@beis.gov.uk
http://www.gov.uk
Our ref: MCB2020/18660
Your ref: ZA97947

8 June 2020

Dear Daniel,

Thank you for your email dated 3 June, on behalf of your constituent, Mr Philip Sergeant of […] CB1 3QJ, regarding amendments to Contracts for Difference (CfD) for low-carbon electricity.

I am grateful to you for sharing Mr Sergeant’s comments. The current consultation on changes to the CfD scheme closed on 29 May 2020. We are currently considering all responses received and will publish the Government response with our decisions in due course.

Changes to the CfD standard terms and conditions are made following engagement with stakeholders and the wider sector. Prior to each new allocation round we publish the final version of the standard terms and conditions applicable to that round.

In Part A of the Government response to the consultation run in 2017 we stated that we will require all dedicated biomass with Combined Heat and Power (CHP) and energy from waste with CHP schemes who are applying for new CfD contracts, to have a minimum 70% overall efficiency, this is confirmed on page 14.

In Part B of the Government response to the consultation run in 2017 we stated that the new lower greenhouse gas threshold of 29 [k]g CO2e/MWh would apply to projects that are offered a contract from the next CfD allocation round, this is confirmed on page 32.

The reason why no mention was made of these standards in our 2020 consultation is because we have already clearly set out that our intention is for these standards to apply for all new contracts, not just those allocated in AR3. As it is already in the contracts it does not need to be consulted on.

The CfD contract is a private law contract between an electricity generator and the Low Carbon Contracts Company Ltd. The greenhouse gas and efficiency limits for biomass were incorporated into the standard terms and do not require secondary legislation.

We are consulting on our intention to remove biomass conversion technologies from the CfD scheme. However, we are not currently looking to make all biomass plants ineligible to compete for CfDs.

Sustainable, low-carbon bioenergy is helping us move to a low-carbon energy mix, increase our energy security, and keep costs down for consumers. We have introduced mandatory sustainability criteria for biomass for heat and power generation. This is to ensure biomass reduces carbon emissions and is sourced sustainably. Generators only receive subsidies for the electricity output which complies with our sustainability criteria. Thank you again for taking the time to write. I hope you and Mr Sergeant find this information helpful.

Yours ever,
RT HON KWASI KWARTENG MP
Minister of State for Business, Energy and Clean Growth

Added note from Philip:
Yes, the CfD document does say 29 kgCO2e/MWh on page 35; as do most DECC and BEIS reports. This is the same as 29 gCO2e/kWh which is the unit most analysts use.

From: Philip Sargent
Sent: 26 May 2020 14:03

To: Daniel Zeichner
Subject: Contracts for Difference (CfDs) for Low Carbon Electricity

Dear Daniel Zeichner,

I was an engineer in DECC/BEIS for 4 years. (We met briefly at a HoL launch event.).

I worked on bioenergy accounting and biomass energy models. I know very well that the zero-carbon accounting currently used in law is not based in any sound economics or physics.

So the carbon footprint limits in this subsidiary legislation is much more important than it looks.

I am contacting you regarding proposed amendments to Contracts for Difference (CfDs) for Low Carbon Electricity. The consultation closes to the public on 29 May, but you will have the opportunity to contact Secretary of State Alok Sharma and Energy Minister Kwasi Kwarteng while the consultation is being reviewed.

In 2018, BEIS announced a 70% minimum efficiency requirement and a stricter greenhouse gas limit for future CfDs for biomass power stations, which were incorporated into rules for 2019 CfD allocation. They stated that without those new rules, CfD subsidies would go to new biomass plants that would increase, not decrease, the carbon intensity of the national grid, i.e. that the safeguards were vital for the UK’s climate goals. As a result, virtually all CfDs awarded in 2019 went to offshore wind, which genuinely helps reduce carbon emissions, and not to biomass plants burning wood from forests.

However, these important changes were only incorporated into the rules for the 2019 CfD auction. The current consultation makes no mention of plans to extend them to future auction rounds. This is very worrying: unless those two changes are reflected in regulations that apply to future CfD awards, we risk spending hundreds of millions of pounds a year in additional subsidies on biomass power plants that emit as much CO2 as ones that burn coal, and that burn wood from biodiverse forests.

The power station currently being built by MGT Teesside is classed as ‘combined heat and power’ even though it is even less efficient than the UK’s largest biomass burner, Drax Power Station; and, like Drax, it is set to burn wood pellets from the Southeastern USA, from a company (Enviva) shown to regularly source wood from the clearcutting of highly biodiverse coastal hardwood forests. Under the standards announced in 2018, these wood pellets would not meet the greenhouse gas limit; it is imperative that the standards are retained so that no further power stations are subsidised to burn these pellets.

Therefore, I would be very grateful if you could urge the Secretary of State to ensure that:

1. Biomass plants are ineligible to compete for CfDs with offshore wind and other renewable technologies in Pot 2;
2. The greenhouse gas threshold and minimum efficiency requirement applied in 2018 will be included into regulations to ensure that both are applied to all future allocation rounds.

Future subsidy awards must go towards the lowest-carbon and cleanest forms of renewable energy only – not to dirty biomass.

Thank you for your time.

Yours sincerely,
Philip Sargent
England, CB1 3QJ, United Kingdom

___________________________
This email was sent by Philip Sargent via Do Gooder, a website that allows people to contact you regarding issues they consider important. In accordance with web protocol RFC 3834 we have set the FROM field of this email to our generic no-reply address at campaigns@good.do, however Philip provided an email address xyzzy@xyzzy.xz which we included in the REPLY-TO field.

Please reply to Philip Sargent at xyzzy@xyzzy.xz.

To learn more about Do Gooder visit www.dogooder.co
To learn more about web protocol RFC 3834 visit: https://tools.ietf.org/html/rfc3834

CHP – what’s it good for?

Well obviously an isolated Combined Heat and Power (CHP) plant makes better use of fuel. But there are several meanings of the word “better”; and much of the advantage depends on what kind of electricity grid the power is connected to.

Isolated CHP

The way to look at a CHP plant in isolation is not to look at what is produced, but to look at what is wasted. Some energy always goes up the flue as hot gas, or is discharged as warm water.So:

• a modern condensing boiler at 83% efficiency (producing hot water) wastes 17% of the energy in the fuel.
• a boiler raising steam (non-condensing) may be 70% efficient and so send 30% of the energy in the fuel up the flue stack.
• a combined-cycle gas-turbine (CCGT) produces electricity at 60% efficiency, thus sending 40% of the energy in the fuel up the stack and into cooling water.
• a good quality CHP plant with a heat:power ratio of 3 may be 60% efficient for heat and 20% efficient for electricity, and so send only 20% of the energy in the fuel to waste.

[Looking at what is wasted avoids doing silly things like adding MWh of heat to MWh of electricity and getting notional ‘overall’ efficiency values which have no real meaning.]

So if you are an isolated paper-making plant which needs steam and electricity, then it makes a lot of sense to install a CHP plant instead of a steam-boiler and in-house CCGT (even if you could get one: most industrial plants have to use something cheaper and less efficient). The industrial plant is then wasting only 20% of the fuel energy, not 40%.

How you calculate the emissions intensity of the electricity generated by a CHP plant depends on what you consider as the alternative heat plant. Suppose we take the example above and consider that instead of a CHP plant, we install a boiler raising steam and buy grid electricity: what grid intensity would the grid have to be to make emissions the same as running the CHP plant?

Realistic CHP

But in industrialised countries there is an electricity grid and the interactions of a CHP plant with that can change everything. It is also not just about minimising fuel use as we need to think about decarbonising too.

In a grid such as the UK’s there is a baseload of nuclear, mostly-there but sometimes intermittent wind, and a load of various thermal fuelled plants: coal, CCGT, OCGT, CHP some of which run all the time. In 2017 the average effective CO2 emissions was about 230 gCO2e/kWh (not counting biomass fuelled electricity) and in 2016 it was 269 gCO2e/kWh.

Since most CHP runs 24 hours a day all year, it is tempting to compare the carbon-intensity electricity from a CHP plant with the average UK grid intensity (all year) and conclude that CHP is becoming a carbonising, rather than a decarbonising technology. This would be incorrect (or at least, not true yet).

Biomethane – future prospects

This article is only about biomethane from food waste – which is one of the most climate-friendly feedstocks for biomethane. A previous post was about biomethane in general and made the point that much of it was probably being produced in the UK at the legal limit of carbon intensity, i.e. not very good at all really.

In summary, biomethane made entirely from food waste in the UK in the future is likely to have a carbon footprint equivalent to an 81% decarbonisation compared with fossil natural gas. This is fine for the UK 2050 target, but inadequate for the Paris Agreement “net zero” target.

[This post was written on 20th Jan. 2018, with an addition on the 21st and a correction on the 22nd.]

Biomethane in England

Today biomethane injected into the gas grid benefits from the common assumption that it is effectively zero-carbon. This is the implication behind many news stories but it is not true. If it is coming from collected food waste then it is better than that: it is negative carbon. This is good news but it won’t always be true.

Biomethane from food waste has an enormous benefit today because if those wastes are not collected for anaerobic digestion (AD) then, in England, they are probably going to landfill. In landfills a worryingly large proportion of the carbon content of the food is lost to the atmosphere as methane – a much more potent greenhouse gas than the carbon dioxide which results when the food is composted or the biomethane is burned.

So the biomethane from food waste is good not because it is intrinsically good, but because the alternatives are very bad. So as these bad alternatives are cleaned up, the carbon advantage of biomethane over fossil natural gas declines. This is not what everyone expects: we expect that things get greener as bad things are stopped, not that something gets worse.

Food waste has already been cleaned up in Scotland, Wales and Northern Ireland. In those places food waste is strongly discouraged from being sent to landfill, so in those places the biomethane from food waste has to largely stand on its own feet and not rely on a bad counterfactual.

Unfortunately there is rather less food waste available to AD than was originally thought (see pages 45-47 and the section on “carbon cost effectiveness” in the renewable heat incentive (RHI) 2016 consultation response impact assessment).

Direct emissions from AD

So what does “standing on its own feet” mean for AD and food waste? It means that we can forget upstream credits: the biomethane has to count all the direct emissions from the AD process against the calorific value of the biomethane with nothing subtracted.

Since there are no credits, there will be a positive emission. So the biomethane will not be zero-carbon, it will not be better than zero-carbon, it will be slightly (we hope) or a lot (we fear) worse than zero-carbon.

So in Table C15 (from the Renewable Heat Incentive (RHI) Impact Analysis) we can’t use that -0.7486 number (= – 748.6 gCO2e/kWh):

so the net emisisons for biomethane from food waste are just those from leakage i.e. +0.032 (= 32 gCO2e/kWh), not -0.604, plus whatever emissions arise from the operation of the AD plant (the parasitic loads) which the impact analysis (IA) neglected to count. The IA also stated that emissions from truck transport of the food waste was “negligible” (footnote 40) and this is calculated at the end of this article.

Emissions of 113 gCO2e/kWh

In Table C15 it might appear that the number -0.113 refers to the overall emissions of a typical AD plant excluding the items which appear in the other columns. This is not the case: 113 gCO2e/kWh is in fact an aspirational goal for upper bound of the process as a whole, being set equal to 90% of the RHI legal maximum emission.

The format of this table is unfortunate in that the meaning of the first column is not made clear. (I misread it myself, and I had a hand in the early work in doing those calculations in the Dept. of Energy and Climate Change.)

Biomethane leakage

It is worth noting that the 32 gCO2e/kWh value for leakage emissions in Table C15 is a calculation and not a measurement. It is simply an assumption that the entire feedstock and digestate transport and digestion process leaks 1.5% of the biomethane produced. This is probably a bit low for the industry as a whole, especially in the UK which has a far less mature AD industry than in much of Europe.

In December 2017, IEA Bioenergy Task 37 produced a detailed report Methane emissions from biogas plants (52 pages) showing leak rates between 0.4% and 5.5% when measured remotely, but up to 13.7% when measured directly. The level of 0.4%-0.6% is current best practice, but unfortunately fairly rare. This report also mentions the usual assumption that parasitic loads are of the order of 10-15%.

 

Biomethane carbon footprint

So in parts of the UK where food waste does not go to landfill, the carbon footprint of the biomethane injected into the gas grid at a 100% food waste plant which leaks 2% of the methane generated (as per page 10 of the NNFCC annex on emissions), and taking into account the 15% parasitic load (which means that only 85% of the biomethane generated is injected), we get an effective biomethane carbon emission of +50 gCO2e/kWh. This is very good (much better than the +202 gCO2e/kWh emissions from maize-fed AD biomethane in the Po valley), and compared to fossil natural gas (260 gCO2e/kWh**) it is a decarbonisation level of 81%. Not zero though.

The obvious way to improve this footprint is to reduce the methane leaks. The footprint can only be reduced to below zero, however, by capturing some of the 40% of the carbon atoms being emitted as CO2 and sequestering them away from the atmosphere.

Composting – the other counterfactual

In my previous post I assumed that composting only produced CO2. Thus I had assumed  that if composting was the counterfactual for food waste, then sending the food waste to AD instead did not imply any “upstream negative emissions” (unlike landfill).

But further research reveals that even well-aerated “closed” municipal composting plants emit some N2O; and N2O is an even more serious greenhouse gas than methane (CH4). Worse, unlike methane, the lifetime in the atmosphere of N2O is nearly as long as CO2.

The 2010 Dutch report reporting the UBA 2008 project says that we should expect between 40 and 100 gN2O to be emitted from each tonne of food waste when composting. This was remarked upon in a later blog which mentioned a more recent number of 55 gN2O/tonne of wet waste. This is 16,390 gCO2e/tonne (or 16.4 gCO2e/kg) using a GWP multiplier of 298x (AR5 Ch.8) for N2O.

If we use:

• that AD converts 75% of carbon atoms into biogas (Prof. Banks)
• that biogas is 60% CH4 and 40% CO2,
• then 1 kg of dried food waste produces 0.45 kg of CH4
• but food waste is very wet so 1 kg of food waste will produce much less than this.

So we get an “upstream negative emissions” N2O footprint of  at most 16.4 gCO2e for 0.45kg of CH4, or 36 gCO2e/kg CH4.

The calorific value of CH4 (LHV) is 50 MJ/kg, so the negative footprint is at most 0.73 gCO2e/MJ or 2.6 gCO2e/kWh. This should be compared to the 50 gCO2e/kWh calculated above. So the N2O contribution is nearly insignificant, especially considering that this is an upper bound and it will be less than half that in reality, say 1 gCO2e/kWh. Since the composting is a counterfactual,it has a negative effect on the biomethane, so the biomethane footprint should perhaps be 49 gCO2e/kWh.

So we can conclude that composting really is very close to emitting pure CO2 and does not afford AD any significant counterfactual upstream negative emissions.

Food waste transport costs

The RHI impact analysis of AD said that transport emissions are negligible, but t would have been equivalently accurate to say that transport emissions should be discounted.

Since the food waste has to be transported anyway, and the distance from collection point to the AD facility would be probably largely similar to that to an incineration, composting or landfill site, then there is no particular advantage or disadvantage for sending the material to AD.

In fact the emissions might be significant but only for long distances. A 10 tonne heavy goods vehicle carrying 2 tonnes emits about 400 gCO2e/tonne-km*, so if the journey is 8km (5 miles) with the truck full one way and empty the other, that makes 3.3 gCO2e/kg of wet waste. This is about one-fifth of the emissions we calculated about for the N2O effect from composting, so it is indeed negligible for short distances.

Trucking food waste 80 km further just to get to an AD plant instead of a composter would have an effect. It would add about 5 gCO2e/kWh to the biomethane footprint: an increase of 10%. But in those circumstances some kind of de-watering process is likely to be done to save fuel costs, so the impact would be lessened.

Resource availability

[Thanks to Richard Lowes for reminding me of this.] We can expect the availability of food waste in the UK to increase somewhat over the next 5 years or so as England starts to act to prevent it going to landfill and as Wales and other regions become more effective at it.

But from 2025 onwards we should expect food waste to be a steadily decreasing resource as we all get better at preventing waste in the first place. Also there will be a steady stream of innovations which can make particular use of specific wastes (usually food processing wastes) which are then more valuable activities than sending it to AD.

Footnote

The numbers in this post give an impression of precision which is misleading. Variations in feedstocks, water levels, transport distances, process operation etc. probably means that any of these numbers might easily be wrong by 30%.

* The 2017 updated emissions for transport are in
https://www.gov.uk/government/publications/greenhouse-gas-reporting-conversion-factors-2017 

** That 260 gCO2e/kWh from fossil natural gas comes from the previous posts on biomethane. But it can’t be right: the 2013 MacKay-Stone report calculates that conventional gas in the UK is 199-207 gCO2e/kWh (chemical energy of the gas). So that would mean that biomethane is not 81% decarbonising, but only 75% decarbonising. LNG would be about 270 gCO2e/kWh so that may have been the comparator used in the previous studies.

Biomethane – Degrees of decarbonisation

Recently a POST note was produced on decarbonising the gas network. An early draft had a mistake in the number quoted for the carbon footprint for biomethane and finding the source of this error turned up a number of interesting issues which some may find disagreeable.

In short, biomethane is probably not as green as it is painted. It has a useful role, especially (possibly only) if some defects are rectified; but the greener it is, the less there is of it. Expansion of biomethane options needs to take account of these intrinsic difficulties. Capture of the vented CO2 may become necessary and profitable.

[This post written 29th Nov. 2017 and updated on 7th Dec. 2017 and 20th Jan.2018]

Conversion Factors for GHG reporting

The POST team used official statistics from BEIS designed for companies reporting their emissions “Government emission conversion factors for greenhouse gas company reporting“. Unfortunately those numbers are not actually used for any real purpose except as one of the inputs to the UK emissions report to the UNFCCC and to track the UK progress against the carbon budget process. So while they should be accurate, nobody gets upset if they are only approximate. Any fallout from inaccuracy would come years  in the future and only on an aggregated basis: nobody’s job is on the line. So I would expect that they are the last set of data to be updated. If we want the best and most recent data we should look elsewhere.

Nevertheless it is instructive to look at those official “conversion factors” as this will set the context for obtaining better numbers elsewhere. This is a bit of a treasure hunt along a chain of official publications and illustrates why it is sometimes so hard to get a simple, correct answer to a simple question. The most recent 2017 reports include spreadsheets of data and a methodology paper. (Reading the spreadsheets makes these all relate to “pure” biomethane, before it has the calorific value corrected to make it identical to natural gas.) The data for biomethane is given in the methodology paper in Table 42 on page 83, and we see a footnote that the number for biomethane Total Lifecycle is 10.11 gCO2e/MJ and actually comes from a different report, from the Department of Transport:

This indicates that we are in trouble. This data is implicitly relevant for biomethane used for road fuel, which is not typical of the biomethane injected into the grid. Indeed there is only one plant supplying biomethane road fuel in the UK but there are now over 80 AD plants in the UK injecting biomethane into the gas grid. Note also that the CH4 and N2O emissions – which increase the value by 10% – are presumably typical of combustion in a road vehicle and not combustion in a boiler. So these too will also be incorrect if we want to know the carbon footprint of injected biomethane.

Road fuel biomethane

The original version of this data is published by Dft “Biofuel statistics: Year 10 (2017 to 2018), report 1” and the relevant part of the spreadsheet is:

Now this is a bit odd. We know there is only one plant putting biomethane into vehicles (Leyland, Lancs.), and we are pretty sure that they are using UK food waste from Widnes, not sugar beet waste imported from Hungary or food waste from Austria. So this is pretty certainly corrupt data. (One has to be suspicious that may be copied from data provided by Czechia or Slovakia.) It is also misquoted: BEIS says 10.00* (plus CH4 and N2O emissions) whereas this is clearly 10.5 gCO2e/MJ (i.e. 37.8 gCO2e/kWh presumably LHV calorific value**).

Now the Leyland filling station uses the green gas certificate mechanism (see footnote #), so it is actually filling trucks http://www.gasvehiclehub.com/ with fossil natural gas. So to find the correct feedstock for DfT to use, they should be using the  AD sites which are issuing those certificates. And indeed it seems to be food waste “by Green Gas Certificates from AD plants that use waste feedstock, such as ReFood in Widnes” http://www.barrowgreengas.co.uk/bgg-in-brief/2016/5/31/gas-to-grid-coming-of-age but they don’t give a complete list of feedstocks or say precisely which plants these are. We could use the BEIS RHI statistics to finds the exact list of feedstocks at Widnes…  or we could look at the 2017 list provided by Green Gas Certificates but this road-fuel data is all beside the main point.

What we want is the mean GHG saving for all the biomethane injection in the UK – not the corrupt data copied by DfT from a single plant or sub-set of plants which may be atypical. So we shouldn’t be using this corrupt data from DfT at all. The master reference document for how this is all calculated is issued by Ofgem.

[It is worth noting that where RHI subsidy is claimed for heat produced by burning biogas, the sustainability regulations apply to the heat (page 8), not the biogas calorific value. So the GHG criteria are significantly stricter for this than they are for biomethane injection.]

UK-wide biomethane injection carbon accounting

We could  use the Sustainable Gas Institute (SGI) report for biomethane where pp70-71 of that report says less than zero up to 450 gCO2e/kWh for biomethane, depending on feedstocks and counterfactuals. This is repeated in the conclusions: “Biomethane CO2  emissions range from -50 to 450 gCO2e/kWh” on page 91. But in the UK, the RHI will be taking a dim view of anything more than 125 gCO2e/kWh as that is the legal maximum for RHI (Renewable Heat Incentive) subsidy.  [125.28gCO2 equivalent per kWh  is 34.8gCO2e/MJ]. So that gives us a more restricted likely range of -50 to +125 gCO2e/kWh.

BEIS don’t seem to be reporting very fully on the RHI feedstocks – each plant has to submit a feedstock report to Ofgem but Ofgem don’t have to publish those very promptly.  We know that the carbon footprint depends very sensitively on the feedstocks used : whether they are wastes or crops, and what would have happened to those wastes if they did not go to AD. So without those detailed reports on each of the 80 plants we have to fall back on the published impact analysis of the injection policy as a whole. Page 64 of that impact analysis is “Greenhouse gas abatement” but it is mostly phrased in the effectiveness of the policy as a whole in MT of CO2, not per kWh of gas, so it is table C15 on page 70 we want:

But (sigh) they have changed the units from g to kg, so these numbers are all 1000x smaller than the ones we usually use: so that is a range from -604 to +145 gCO2e/kWh for food waste and crops respectively. [ See later post about the 113 gCO2e/kWh number. ]

Profit-seeking operators will want to use crops, whereas BEIS wants them to use slurries and food waste. We know this from the studies which re-evaluated the level of subsidy and implicitly showed that crops are more profitable: “RHI biomethane injection to grid tariff review“. (Actually crops are really worse than +145 because they have upstream emissions from fertiliser, esp. N2O emissions, as well as cultivation and harvesting carbon costs.) So we can assume that many AD operators will be working at the maximum legal limit, which is the profit-optimal position, which is 125.28 gCO2e/kWh.

When the published revised RHI regulations become effective early in 2018 (in parliament now, very delayed by Brexit, election etc.) new AD plants will get RHI subsidy on biomethane from crops only up to the amount of gas produced from wastes and residues, i.e. payments are 50:50 for gas from wastes and crops.

If we assume that wastes give -604 gCO2e/kWh and crops give 125 gCO2e/kWh then this means that new AD plants will be operating at -240 gCO2e/kWh (the average of the two extremes) which looks excellent… But that calculation assumes that all the wastes are food wastes and assumes the counterfactual of all food waste going to landfill if it doesn’t go to AD. Landfill is climatically horrible because (as assumed in the impact analysis) it leads to 35.3% of the methane escaping to atmosphere. But even so that -230 gCO2e/kWh number is only valid in England where landfill is an option. In Wales, Scotland and Northern Ireland food waste is not allowed to go to landfill: the alternative to AD is incineration or composting, neither of which lead to methane emissions. So outside England (and in England after landfill for food waste becomes illegal, probably before 2030) we have to  assume that the counterfactual for food waste is municipal composting, not landfill, which would give an upstream correction of zero (biogenic CO2 is counted as zero under current Renewable Energy Directive rules). Then the emissions from AD using food waste would be just column 1 plus column 2 from the table above: 113 + 32 = 145 gCO2e/kWh… oh dear. That’s above the legal limit even without using any crops. [ See later post about the 113 gCO2e/kWh number: it is not what it appears to be and my calculation of 145 gCO2e/kWh is wrong.  I have now done a corrected calculation of the carbon footprint of biomethane from food wastes] Still, farmyard manures for AD look good – until farmyards clean up their current handling of manures and reduce the methane emissions anyway without AD.

If that is the situation for the future, what about the existing 80 sites which were mostly built before the carbon accounting rules came in (“sustainability rules”)? They will be using a lot of crops (in general) to maximise income and so will be rather worse than the current legal limit.

In fact ADBA have published data on farm-based dedicated agricultural AD plants: all 273 of them (which are mostly biogas CHP) in their Winter 2017 News. These plants use only agricultural feedstocks: 55% from dedicated energy crops and 45% farm wastes (page 10). These percentages are presumably by tonnage not by biogas-productivity which means that 55% is a lower bound for source of the gas from crops. This means that the existing AD industry probably has a negative GHG intensity because of the benefit (today) from the conventional (bad) slurry disposal outweighs the emissions from the food crops. (But the on-farm food waste would not otherwise be going to landfill, so the -749 gCO2e/kWh number would not be appropriate for that.)

Annex 3 (pp96 onwards) of the published impact analysis  makes some ad hoc calculations to look at indicative sensitivities with no basis in evidence:

“For our lower bound (optimistic) estimate of carbon cost-effectiveness we have assumed that emissions from AD are 20% lower than the sustainability criteria limit where agricultural feedstocks are used; and 90% lower than the sustainability criteria limit where food waste is used.”

And then say they’d like help improving this

“Our approach to estimating carbon savings from AD serves to provide an indicative estimate of £/tCO2e saved based on the emissions limits under the UK’s biomass sustainability criteria. Insofar as the emissions impact of AD differs from the sustainability criteria limits we welcome further evidence which may help to better understand the carbon cost effectiveness of AD.”

But it’s better than that. Page 97 quotes several bits of evidence for emissions from food waste in landfill, from which they conclude:

we consider that, for the purposes of our analysis, a sensible assumption is that food waste AD produces an upstream emissions abatement effect of between -450 and -900gCO2e/kWh of energy produced, with a central assumption of –790gCO2e/kWh.

[ I don’t know why table C15 has 748.6 instead of 790.0 but I regret to observe that editing consistency is not as reliable as it should be in any of these reports.]

So Fig. A3.1 of the impact analysis summarises all this in terms of the cost of carbon abatement at the proposed subsidy levels. [ For reference, a figure of more than £100/tCO2e is generally considered “rather expensive” for UK carbon abatement policies and is only justifiable if the policy is aggressively encouraging deeper decarbonisation and there is a realistic downwards trajectory. ]

which basically says that AD for anything but food waste in England for the next few years is not a good use of public funds (for an unrealistically small methane leak## from the biogas upgrading unit of 0.5%)  and implicitly that public funds should not be spent on AD plants using any crops.

Some biogas upgrading units emit 3% of methane to atmosphere by design so would struggle to be better than £100/tCO2e even for 100% manures and slurries. The  “RHI biomethane injection to grid tariff review” lists several options that could be used in policy revisions to improve this in future.

Comparison with biomethane in Europe

What about EU data?

Page 134 (injected biomethane) of the Joint Research Centre publication  gives -103 to +58 g/MJ i.e. -371 to 209 gCO2e/kWh. Table 102 on page 136 gives the breakdown and we can see the big emissions contribution made by fertilizers (N2O emissions) and cultivation when maize is used. This confirms that the UK data from the BEIS publications is probably in the right range, and also confirms that the carbon footprint of biomethane depends very significantly on the individual plant and the feedstock it uses.

Recent work in the alluvial Po river valley in northern Italy proposes a system of double-cropping providing biogas at -335 to 25 gCO2e/kWh (page 6 of their February 2017 summary report). (This also confirms that 100% maize-fed AD produces 202 gCO2e/kWh which is only a 22% decarbonisation compared with natural gas.) However whether this sequential cropping method is applicable to the very different climate in the UK and to soils with nitrogen-limits  requires UK-specific studies. There is as yet no published detail on these Italian numbers yet so we don’t know if they include the effect of methane emissions from the upgrading system or not.

So there is no simple answer: it depends on farmers’ current handling of slurry and whether a new AD plant properly manages the liquid digestate or not (which continues to give off methane for weeks). We don’t even have any numbers for “current best practice”.

What to do to improve the data

Biomethane injected into the gas grid comes from

• Sewage
• Agricultural manures
• Agricultural wastes: unsellable food and crop residues
• Food industry processing wastes
• Food wastes from restaurants and hotels

Note that there is a concern about food wastes from households because of potential  contamination from collections from the public and the implications this has for disposal of the digestates. About half of all food waste is already collected, so the public campaigns to improve food collection in England will make a useful but not dramatic improvement. The November 2017 report of the ADBA covers this in some detail (pages 36-44).

In gas volume terms, not much of it comes from manures either: although a lot of manures are used in tonnage terms, manure doesn’t produce as much useful gas as food waste: only about a tenth as much.

There are some aspects which mean that biomethane is somewhat better in GHG emissions terms than the current accounting system allows. The use of a break crop or cover crop in AD improves the productivity of later crops. The solid digestate is a soil improver and creates better crops in the next season. The liquid digestate is a nitrogenous fertiliser and offsets the energy required to make artificial fertiliser. (But the digestate does not offset the N2O emissions as these are intrinsic to all fertilisers and depend largely on the timing of their application.) Unfortunately there are limits to the amount of nitrogen that is allowed to be applied in many UK farming areas so this benefit is not available to some AD plants.

The most profitable feedstocks tend to be crops: the ones that emit most CO2e during cultivation (including especially N2O from fertiliser application). This is especially true now that AD is established because gate fees for wastes have reduced. So a profit-seeking operator will choose feedstocks that just meet the legal limit: just below 34.8 gCO2e/MJ. So one can expect that that will be the actual emission, not just the maximum emission, from AD. But we will have to wait for audited reports from Ofgem before we know for sure.

Note that nearly all AD plants injecting biomethane also have CHP biogas engines on site providing heat for the digesters and power for the handling equipment. These engines get FITS subsidies for the power and RHI subsidies for the heat, but there is as yet no planned requirement to use 50:50 wastes feedstocks for FITS subsidies. Thus one could expect slightly more crops to be used than expected and all that biogas would be accounted for under FITS. A full carbon accounting for biomethane injection should ideally take these CHP engines into account.

Source of the 34.8 gCO2e/MJ number

This is documented in a very unusual one-page note written by DECC in 2014 when it came very close to shutting down the entire nascent AD industry by accident. The number 34.8 gCO2e/MJ comes from EU publications and is a 60% decarbonisation of the EU average “fossil fuel counterfactual” (FFC) for heat. So those MJ are MJ of heat after the FFC has been burned in a boiler which is assumed to be only 77% efficient (EU standard number). Representations from the UK AD industry made it clear that this would be an almost impossible target to hit in 2015, so DECC applied the number 34.8 gCO2e/MJ to the calorific value of the biomethane instead, which was an effective relaxation of the target by 30% (i.e. 100/77).

For biomethane injection the real counterfactual is the calorific value of the methane content of the natural gas as the biomethane substitutes for that directly. So this GHG limit should really be changed (reduced by 17%) to 104 gCO2e/kWh [28.9 gCO2e/MJ] for a 60% decarbonisation.

[update 23 Sept.2021: the new Green Gas policy which replaces the RHI is requiring a 24 gCO2e/MJ but not admitting to any previous mistakes.]

If one aimed at a 70% decarbonisation (as the DECC note suggests) which would be approximately appropriate for the 5th Carbon budget period 2027-2032 then this should be 78 gCO2e/kWh [21.7 gCO2e/MJ]. Meeting this in real AD plants will require great care and very careful planning.

The current level of 34.8 gCO2e/MJ corresponds to a decarbonisation of 52%. So if one could substitute half the gas in the grid by this biomethane then this would still only be an overall gas grid decarbonisation of 26%.

These calculations use the same GHG footprint of natural gas – including both upstream emissions and combustion – as the DECC note in 2014. In reality there would be some slight changes in future as the proportion of imported gas changes and as Qatari LNG starts to use decarbonised electricity in the LNG production process.

Solutions for the future

What can we expect AD to be doing in 2025 when we can expect the next wave of policies to be coming into effect?

• In 2025 we could expect good AD methane monitoring, better practices for accidental leaks and methane emissions from post-AD digestate, proper chemical (amine) systems for biogas upgrading which do not emit methane and better accounting for use of digestate to displace artificial fertiliser: all of which make biomethane better in terms of CO2e GHG footprint.
• However in 2025 we can also expect that agricultural practices, food waste in landfill and farm slurry handling to also be much improved. All of which destroy the “bad counterfactual” cases for biogas and so make AD appear worse than it does now. Much worse.

The relatively pure biogenic CO2 created by AD is nearly all currently vented. This is assumed under current accounting conventions to have zero climate impact. If this CO2 were collected and sequestered, then a very substantial negative carbon emission would accrue. If there were a market for negative emissions then this would become a very substantial extra income for the AD operator and would also completely transform the GHG footprint of the industry.

Footnote on trading and perverse incentives

When there is such great variation in the carbon intensity of biomethane from different sources it makes it very difficult to construct rational biomethane trading or certificates markets. Indeed premature formation of such markets can freeze carbon intensities at current (poor) levels and create a profound disincentive on all actors in cleaning up the industry – despite that being in the best interests of the industry as a whole.

Footnote on updates

In January 2018 I discovered a misreading of the data in the impact analysis and extended the work to cover the “negligible” emissions from transporting food waste. This is covered in a later article.

*  I had to change the formatting: the downloaded spreadsheet only give zero places of decimals, so I had to change the formatting to show the resolution we need.

** LHV = Lower Heating Value: the calorific value from burning the fuel assuming that all the water vapour is not condensed back to liquid and the latent heat recovered. Ofgem use “net” calorific value in their guidance (page 22)  because they were working from the RHI official regulation on this matter (Schedule 2A) which is copied from EU RED regulations and uses net calorific value as standard for all fuels (para 4.c). “Net” is the same thing as “lower” for dry fuels. However this is not really the right thing to do if the gas is being burned in a condensing boiler for which the gross (i.e. higher) calorific value is appropriate. Using “net” rather than “gross” means that the GHG emission limit is stricter by 11% since that is the ratio between the two numbers for pure methane; see https://en.wikipedia.org/wiki/Heat_of_combustion . [I would appreciate someone checking my arithmetic on this issue.]

# The Green Gas Certificates scheme issues certificates for biomethane injected into the UK gas grid which has already been subsidised by the RHI injection scheme: see the page on “additionality” on their website. The level of subsidy in the RHI is intended to exactly compensate for the difference in cost of producing injectable biomethane compared with fossil natural gas, so in principle “green gas” should cost the purchaser no more than the ordinary cost of gas (plus the overhead for operating the certificates mechanism). If any cash finds its was back to the AD operator then that is an argument for reducing the RHI subsidy level in accordance with the EU State Aid Rules and Treasury guidelines for this type of policy.

## IEA Bioenergy Task 37 published a comprehensive study on methane leaks in December 2017 Methane Emissions from Biogas Plants . This report covers a great deal more than leaks – it reviews all the life cycle assessments of greenhouse gas emissions.