Comparing blue hydrogen with natural gas emissions

Hydrogen is in the news a lot lately, because of the publication of the UK government hydrogen strategy document. I was alarmed to find a recent article [1] saying that blue hydrogen (made from methane gas with carbon capture) is hardly better than grey hydrogen (made from methane without the carbon capture). The authors are Robert Howarth from Cornell University and Mark Jacobson from Stanford University). I have recently been saying that blue hydrogen is at best about a third the emissions from methane gas, which I think is bad enough - but this article suggests even that is naively optimistic. 

Obviously there are some vested interests trying to paint blue hydrogen as white as possible (pun intended). So I decided to dig deeper into the assumptions behind the analysis. There are three main sets of assumptions involved - about the upstream emissions, about the capture rates and about the source of energy used in the hydrogen conversion plant. Plus you can choose to use 100 year global warming potentials for methane or 20 year ones. Given the climate emergency, the 20 year timeframe seems more appropriate. Also I have chosen to compare blue hydrogen with natural gas, which is the fuel it would mostly replace. Depending on the assumptions you choose, you can make blue hydrogen decidedly worse or significantly better. I can only get it down to 1/3 the emissions of natural gas using the 100 year time frame, not over 20 years. 

Emissions from natural gas and blue hydrogen with the original assumptions in [1]. NG is natural gas, Blue is blue hydrogen, 100 years/20 years means using the 100/20 year global warming potential for methane 

The chart above compares the global warming potential (GWP) for natural gas and blue hydrogen using the assumptions in [1] under both 20 year and 100 year time frames. Using the 100 year GWP (which is more usual), natural gas and blue hydrogen come out about the same, whereas over 20 years the hydrogen is 24% worse. This is because methane has very strong climate heating effects but does not last very long. Extracting and processing methane always involves some level of leakage, though how much is highly controversial. There is a strong argument that targeting methane now would be a good way to make a big difference in a short space of time [2] and from that point of view, it makes no sense at all to be planning to extract (and leak) more of it. 

Upstream assumptions: reduced leakage and cleaner energy

The main upstream emissions (because of the high GWP) are due to methane leakage. The article has made an assumption of 3.5%. Not surprisingly, the actual level of leakage varies from one gas field to another and it is difficult to measure. 3.5% is not an unreasonable estimate but one could be optimistic and assume that this will decrease. One study from Canada suggests that 2% is a better estimate for that region and some fields were more like 1% [3]. So let's assume that 2% is achievable.

Another source of upstream emissions is due to energy use in extracting, processing and transporting the methane to the hydrogen conversion plant. Howarth and Jacobson have chosen a source which estimates this energy as equivalent to 7.5% of the methane delivered, and assumes that this has come entirely from methane. One could imagine that this process could be made more efficient, and perhaps some of it uses cleaner energy sources. I have assumed a lower limit for this of 2%. However this change makes much less impact than the leakage rate.

With lower leakage level and cleaner methane processing, the chart looks like this. Emissions for natural gas and hydrogen are all substantially reduced - down from a worst case of about 500 down to 320 gCO2/kWh. However blue hydrogen is still slightly worse over 20 years.

Emissions with reduced leakage and cleaner upstream processes for methane

Carbon capture rates 90% all the way

Howarth and Jacobson base their carbon capture rates on what is typical in hydrogen factories at the moment. This may not be entirely fair, though, as there are not many around and if this industry does scale up one can hope they will improve. There are three stages where carbon capture is possible - from the chemical reaction (methane + water -> hydrogen + CO2), from the heating required to make the reaction go, and from energy use in the carbon capture process itself. Howarth and Jacobson assume 85% carbon capture for the chemical process, which is fair, though not great. However, they have used a lower capture rate for the heating process: just 65% (which they justify with examples of actual installations). Furthermore, they assume that energy for the capture process also comes from methane and with no carbon capture at all. Whether or not this is current practice, it may not be future practice. It is generally agreed that 90% carbon capture is achievable, so I have applied alternative assumptions for 90% capture at all three stages.

Emissions with cleaner upstream assumptions as above and 90% carbon capture rates

With these assumptions on top of the cleaner upstream processing, we find that blue hydrogen is substantially lower in GHG emissions than natural gas (which has not changed) - but not nearly enough to be useful. Over 20 years the blue hydrogen is only 22% better.

Alternative energy for carbon capture

Methane emissions are the main cause of the difference, now. For each kWh of hydrogen (gross heating value) you need 1.4 kWh methane for the chemistry and the heat, plus another 0.3 kWh energy for the carbon capture. This is one reason why hydrogen made this way can never be as cheap as natural gas. You need 40% more gas just to make it, even ignoring the carbon capture.

I find it hard to imagine using anything other than methane to supply the reaction heat, because it is bound to be cheaper than other energy sources. Also given that they will already have carbon capture equipment on site, it would not be hard to apply this to the flues for the heating furnaces. Capturing carbon from flue gases is harder because there is a mixture of gases there whereas the carbon from the chemical reaction is more pure. But we can and do carbon capture for power stations and this is the same problem. So I have not substituted the energy for the heating stage. 

I am not so sure about the energy for carbon capture. Some of this will be heat too. However, with a stretch of the imagination one could assume that this will be energy from another source which is zero carbon. 

This helps a bit more. Blue hydrogen is now only one third the emissions of natural gas, if you assume the 100 year GWP. Over 20 years it is two thirds the emissions - nowhere near good enough.

The Future Energy Scenarios suggest biomass gasification to offset the emissions - plausible but not proven

National Grid's latest future energy scenarios recognise that blue hydrogen is not zero carbon. However they suggest that a proportion of hydrogen can be made from BECCS (Biomass energy with carbon capture and storage) [4]. Since that carbon is from a plant source it counts as negative. Their 'system transformation' scenario uses a mix of hydrogen sources including 10% from biomass to offset the blue. This works using 100 year GWP, but in the shorter term we would need an awful lot more BECCS than that, to offset the intense warming from extra methane leakage.

In any case hydrogen this way is hypothetical at the moment. There are lots of possibilities but, to my knowledge, no actual pilot programmes. Drax has a pilot power station using biomass and carbon capture but that produces power, not hydrogen.

The National Grid assumptions are broadly compatible with the above. Their estimate of emissions from grey hydrogen is a little lower than Howard and Jacobson; reducing the leakage rate to 2% as I have done is enough to correct for this. For blue hydrogen they seem to assume slightly lower capture rates than my charitable assumptions above.

Conclusions

Arguably Howarth and Jacobson's analysis was overly pessimistic. They used data from current practice and it is certainly possible to make improvements. However, even with kinder assumptions, there is no way that hydrogen production can be zero emissions within a useful time frame. Over the next few decades, while the leaked methane will be having its effect, blue hydrogen is not much of an improvement than sticking with natural gas. Capturing the carbon and using clean energy in the production process is not enough to offset the increased upstream emissions - which are due to the need to extract 40% more methane in order to achieve the same energy value. Offsetting these with BECCS, as National Grid propose, is not effective over a 20 year time frame.

My conclusion is that producing blue hydrogen production at scale is going to have the opposite effect that we are trying to achieve. I am more convinced than ever that we have to stick with cleaner methods such as electrolysis. This means we will have less of it and we must use it wisely.

[1] How Green is Blue Hydrogen? (Robert W Howarth and Mark Z Jacobson in Energy, Science and Engineering) July 2021

[2] Reduce methane or face climate catastrophe, scientists warn (Guardian) August 2021

[3] Methane emissions from upstream oil and gas production in Canada are underestimated (Katlyn MacKay, Martin Lavoie, Evelise Bourlon, Emmaline Atherton, Elizabeth O’Connell, Jennifer Baillie, Chelsea Fougère & David Risk, Scientific Reports) April 2021

[4] Future Energy Scenarios 2021 (National Grid) 2021