Prof Jason Lowe is Head of Knowledge Integration and Mitigation Advice at the UK’s Met Office and lead scientist for the government-funded AVOID2 research programme, which provided impartial advice on the feasibility of pathways designed to avoid “dangerous” climate change.
A question that I am frequently asked, as attention focuses on the consequences of CoP21 last December, is whether limiting warming to 2C above pre-industrial levels is practically feasible?
When we consider the feasibility of limiting global average warming to below 2C, several aspects must be accounted for.
The first is the physical climate’s contribution to feasibility. This determines, for instance, how much warming occurs for a given amount of carbon dioxide emissions, and the rate of temperature change if emissions are reduced or increased. To address this part of the problem, we can make use of earth system models. These contain our best understanding of how the large-scale physical climate system works.
The second aspect of feasibility relates to the technical and economic achievability of making emission reductions. Will technology be available on a large enough scale in order to satisfy energy needs whilst producing lower greenhouse gas emissions? Will such technology be affordable? Unlike for our earth system models, we can’t rely on the laws of physics. But we can represent our current understanding of the way the energy systems and economic aspects operate, including taking a view on how quickly technologies will develop. This happens within energy system and Integrated Assessment Models (IAMs).
The third aspect of feasibility concerns the role of political choices and social barriers. Some of these can be accounted for by the choice of experimental design applied to our computer models of climate, economics and the energy system, such as setting the time when emissions peak or fossil fuel emissions approach a given level.
There are also, clearly, links between the different aspects of feasibility, such as between climate and technological feasibility or between political and economic feasibility, that are sometimes included in investigations of the future, at least to some extent.
A large number of studies, using a range of different IAMs, find cost effective pathways that give at least a 50% or 66% chance of limiting warming to below 2C above pre-industrial levels. This article is focused on one particular feature of these scenarios of future energy use and emissions – that there is a reliance on a technology that combines biofuels with carbon capture and storage (referred to as BECCS). It provides energy, but also removes carbon from the atmosphere by first locking it in plant matter, then capturing it and storing it in suitable long-term storage reservoirs, such as disused oil or gas wells.
We need to answer several key questions. First, how much BECCS is required? Second, is the amount of BECCS required viable without causing unacceptable side effects? And third, if BECCS turns out to not be available in the required amounts, what does it mean for future climate and the ability to limit warming?
How much BECCS is needed?
One of the key results of the IPCC 5th assessment was to highlight the almost linear relationship between cumulative carbon emissions (or global carbon budgets) and peak global average warming level. The use of BECCS allows us to exceed the cumulative emission budgets that would otherwise correspond to a given level of peak warming. We can think of this as going into debt with our atmospheric carbon budget and then paying it back later through the enhanced removal of carbon dioxide.
As part of the IPCC 5th assessment, a database of emissions scenarios from IAMs set up by different researchers around the world was assembled and is available to the research community. These simulations provide a view of the features of different emission pathways.
Of these scenarios, around 500 are ideal for looking across the full 21st century and accounting for multiple gases. Around 200 limit median warming in 2100 to below 2C. We have looked at this experiment set and made an estimate of the amount of carbon removed by BECCS in these experiments during the 21st century.
The figure below shows a cumulative distribution function for different 21st century removal amounts for the full scenario set, and those scenarios that limit warming to below 2°C above pre-industrial levels.
Focusing on the sub-set of 2C scenarios, the median amount of carbon removal is around 170 GtC (1Gt = 1 gigatonne = 1 billion tonnes) or 630 GtCO2, and the 90th percentile of the distribution is 320 GtC or 1200 GtCO2.
To put this into context, the IPCC fifth assessment estimates the emissions of CO2 from pre-industrial time to 2011 as being around 511 GtC, or 1900 GtCO2. So in the more extreme BECCS scenarios, there is a demand to take out of the atmosphere between now and the year 2100 around two thirds of the total amount carbon released up to 2011.
In many of the scenarios considered, late in the simulations, for each year, the amount of carbon dioxide removal becomes greater than the human-driven emissions of carbon from other sources so that a net negative emission is achieved. Typically these scenarios have a large amount of BECCS.
Although in the minority, there are scenarios in the database that can limit warming to below the 2C level without any BECCS. However, many of these tend to peak the global emissions quite early, and before present day. This is not some type of climate scientists revisionism, it reflects that many of the experiments were designed some years ago and the urgent need for more up to date scenarios that reflect recent emissions and changes in the energy mix in the real world.
Is the amount of BECCS viable?
BECCS is an emerging technology but there are several estimates in the literature of the capacity for BECCS over coming decades. There are limitations on the amount of biofuels and the efficiency of extracting energy from them; this includes uncertainties around the amount of land and water availability, and crop yields. There are uncertainties in rate at which Carbon Capture and Storage (CCS) plants can be deployed, the energy penalty imposed on power production by the inclusion of CCS, storage space for the CO2 and leakage rates.
Work in the AVOID2 programme led by Nem Vaughan and Clair Gough searched the literature and found that many IAMs set a maximum limit on energy from biomass (for example of 200 EJ per year), whilst others incorporate more detailed land use modelling. Many IAM scenarios assume that BECCS utilises dedicated rain-fed bioenergy crops grown on surplus agricultural land, assuming medium yields and the use of crop and waste residues. This seeks to circumvent issues of competition with food production and other land uses, but is strongly dependent on the underlying socio-economic assumptions. In future work, there is also a need to be clearer on the potential use and relative merits of perennial compared to annual biofuel crops.
The global potential for negative emissions has been estimated to be between 0 and 10 GtCO2 per year in 2050 and between 0 and 20 GtCO2 per year in 2100. If BECCS starts in 2020, the maximum values equate to around 900 GtCO2 (245 GtC) removed by 2100. The authors conclude that lower amounts could result from weak or no climate policy, lack of social acceptability, and/or failure of the BECCS system to deliver net negative emissions.
In an attempt to move beyond the current literature, the researchers organised an expert elicitation study to understand in more detail some of the barriers to large-scale deployment. The work highlights that the scale at which BECCS is assumed to be deployed in many scenarios is extremely ambitious. Assumptions representing a more modest realisation of BECCS might be more realistic and consequently, better represent BECCS as a feasible climate change mitigation option. The timescale at which BECCS is assumed to be deployed at very large scales is equally optimistic in the models. Assumptions about the rate of deployment, which should be based on precedented technology uptake rates, depend on the rate at which storage can be identified and utilised and at which infrastructure, governance and policy frameworks can be put in place. Furthermore, it will affect the cost of deployment and may influence social acceptance and the societal responses.
Embedded in figures for global net negative emissions is a rich and diverse spatial and regional heterogeneity, which is represented to different levels between and within models, and applies across many assumptions, including storage availability, crop yields, political and social context, policy frameworks that must work globally and locally. The influence of future local and regional climate change could be significant and impact on many elements of the BECCS chain, including the production of biomass and social acceptability. The complexity of systems involved in BECCS approaches should not be underestimated, and is characteristic across the component systems and integrating between technologies, between different actor networks and supply chains.
There remains a gap between modelled or theoretical perspectives and real world experience of building and demonstrating the BECCS process as a whole, and across the supply chain. Despite the heavy reliance of future scenarios on BECCS, the technology is only entering the demonstration phase. There is a large scale demonstration project in Decatur, Illinois, USA and around 15 other pilot scale BECCS plants globally.
A novel piece of work in AVOID2, led by Andy Wiltshire, looks further at the planetary constraints on the biofuel aspects using a new earth system model experiment. The experiment uses the well-known HadGEM2ES model, which can represent both regional climate change and changes in regional vegetation, along with feedbacks between the local land surface response and the local and large-scale climate.
The first key finding is that the land use emissions embedded in BECCS scenarios can be large and may offset some of the potential benefits, especially in the first decades. These need to be included in emission pathways. Looking at several scenarios the work concluded that the gross negative emissions from BECCS are unlikely to exceed 640GtC (2346 GtCO2) over the 21st century. However, when land use change is included the net maximum contribution is greatly reduced in the experiments to 130 GtC (476 GtCO2). The highest gross emissions come from a scenario assuming rapid expansion of bioenergy crops from 2020 in the highly productive tropics covering 18% of the land surface (excluding Greenland and the Antarctic) in 2100. The highest net emission comes from a scenario assuming available abandoned agricultural land is put into production, reaching a maximum of 5% of land cover in 2100. In the scenarios considered here, the effect of regional climate change on crop productivity is included and the biofuel yields increase under climate change, but with important regional variation.
In the absence of limits on CCS, a large constraint on BECCS was found in AVOID research to be the amount of land allocated to bioenergy crops and the rate of deployment. Competition for land for food production is a key uncertainty. Other major constraints on the amount of BECCS achievable are the harvest and sequestration efficiencies. A further key finding of the research was that the other biophysical effects may be very important, such as the change in local surface albedo and the amount of sunlight reflected back into space as the land use changes. Locally, this can either add to or offset the cooling potential of BECCS and this effect is often not included in considerations of biofuels.
We are only starting to quantify these effects in the context of BECCS and due diligence to better understand the capacity and side effects remains to be done. Whilst this earth system model study is an exciting contribution we must remember that it uses only one model and has many assumptions, for instance by focusing on annual biofuel crops. It does not provide a robust upper limit for BECCS but it does add a piece of evidence that suggests some of the larger amounts used in scenarios may not be so easily attainable and relying on this to limit warming to below 2°C is clearly a risk.
What would be the consequence of not having access to BECCS?
We can consider this question first in the context of what would happen in the world of earth system models and IAMs. First, let’s assume that the carbon in the scenarios that is currently removed by BECCS is instead released into the atmosphere. This is equivalent to assuming the energy requirement from BECCS is supplied by carbon neutral sources, such as renewable energy supplies or fossil fuels with perfect CCS.
A rough estimate of the amount of extra warming can be made using the published estimates of the relationship between cumulative carbon emissions and peak warming from the IPCC assessment, called the Transient Climate Response to cumulative carbon Emissions (TCRE). This is the warming per GtC, and has a range in the 5th assessment report of 0.8 to 2.5C per 1000 GtC. Combining this with the 50th and 90th percentiles of BECCS requirement seen in the 2C scenarios based on the IPCC database suggests an additional warming of 0.13 to 0.8C. However, if the BECCS was instead replaced by fossil fuel energy sources without CCS, then the extra warming could be considerably greater than these amounts. This is a very simple estimate and doesn’t account for more complex feedbacks, such as any changes in surface albedo that occur as land use is switched to growing biofuels.
The alternative way to address the question of understanding the consequences of having no BECCS is to say that a “well below 2C” warming limit is still required, and to ask if it is feasible to achieve it without BECCS.
Projects have examined this question, including EU AMPERE and AVOID2, and it is considered in the IPCC 5th assessment using experiments in which particular technologies are not available. The absence of BECCS does not stop a number of IAMs from limiting warming to below 2C with a 50% or higher chance, but it does tend to put up the mitigation costs, as other measures are needed and this typically manifests as a higher rate of fossil fuel decarbonisation. This obviously also carries a risk about whether the other power generation or demand reduction technologies are really able to take up the slack from not having sufficient BECCS. It also raises the need to consider other mechanisms of artificially removing carbon dioxide from the atmosphere alongside BECCS.
Thus, the evidence from modelling suggests it is not impossible to limit warming to below 2C without BECCS but it does again illustrate the scale of the challenge of achieving a limit on warming of “well below 2C”. Without early action to reduce emissions, the available evidence suggests the world will become increasingly reliant on the existence of artificial carbon dioxide removal methods such as BECCS.
Of course, these estimates are from the idealised world of models. What would happen in the real world if BECCS can’t be deployed on a large-scale?
The answer is we don’t know yet. The world could choose to relax the 2C limit a little to give a bigger carbon budget, or society may rise to the challenge of exceeding past rates of low carbon technology roll-out.
Alternatively, the answer may sit between the limiting cases of the thought experiments presented above. What we can say with confidence from the available evidence, is that it is possible to reduce the risks of exceeding potentially dangerous levels of future climate change if sizeable emission reductions take place sooner rather than later. Doing this reduces the risk of relying on either BECCS or other technologies being available on very large-scales. Furthermore, it is vital to also keep in mind that the impacts to society from a range of climate effects and extreme weather are not eliminated even in the most ambitious emission reduction scenarios and there remains a need to better understand how to manage these continuing risks on local and regional scales alongside gaining a better understanding of mitigation technologies, including BECCS.
Do we need BECCS to avoid dangerous climate change? A guest post from Dr Jason Lowe @Metoffice