Developed by contractors on behalf of the Environment Agency (EA), WRATE uses the life cycle assessment (LCA) methodology to evaluate the environmental impacts of waste management systems. Whether it’s recycling steel cans or treating residual waste through MBT, WRATE is the officially sanctioned system that will tell you the benefits.
However, as someone who has become all too familiar with this tool, having used it professionally since it was first released, I’m worried by the status it is given. WRATE analyses are often requested by government bodies, and it continues to be widely used to appraise policy options in the procurement of waste services for local authorities. More recently, the Green Investment Bank relied on outputs from WRATE to appraise the “green impact” of the projects it was considering funding within the waste sector. Perhaps they expect that the tool will contain reliable assumptions that allow different proposals to be weighed on a consistent basis. Unfortunately, they’re wrong.
Does my back end look big in this?
WRATE’s main selling point – and a big part of its development cost – is its sophisticated ‘front end’: a user interface aimed at non-experts. Tool developers Golder Associates claim it is “designed for waste managers by waste managers”. Yet in practice WRATE is mainly used by professional modellers, who have little use for the interface and care mainly about the data and the underlying assumptions in the ‘back end’ of the tool. Unfortunately these have not been lavished with the same investment as the under-used front end…
WRATE was first issued in June 2006, accompanied by a commitment from the EA to provide updates “during the lifetime of the software”. However, there has been only one major software update since the launch, seemingly prompted by some revisions to the ecoinvent database used in the tool. It took three years for these to be incorporated in WRATE V2, which was finally issued to users in 2010. Now much of the standard data in WRATE needs substantial revision – several key datasets date back to 2005 or earlier, and a great deal has changed since then.
Unjustified and ancient
A key element of LCA is the ‘characterisation factors’ that are applied to emissions from facilities. One widely-known example is the need to scale up the climate change impacts of gases such as methane to allow greenhouse gas (GHG) impacts to be expressed by a single metric, typically their carbon dioxide equivalent. WRATE continues to use the multipliers contained within the 1996 version of the IPCC guidance even though this was revised in 2006. A further substantial revision is expected at the beginning of 2014. If current proposals are taken forward, the revisions could see the climate change impact of methane emissions increase again, to exceed the value currently in WRATE by over 60%. This is important: WRATE looks set to disagree significantly with the latest climate science.
WRATE also helpfully contains over 50 other characterisation factors which aim to evaluate environmental impacts ranging from water eutrophication to soil toxicity. However many of these are also now outdated – for example, since 2006, there have been significant changes in the assessment of toxicity in LCA to improve the overall robustness of the results. These aren’t the only problems.
- Major changes were again made to the ecoinvent database earlier this year, but have not been carried through into WRATE.
- The landfill model in WRATE, largely unchanged since 2005, is now out of sync with other models (MELMod and GasSim) used by UK government agencies to estimate landfill emissions.
- Issues with WRATE’s treatment of biostabilised wastes were identified by some in the user community shortly after the launch of the tool and eventually recognised by the EA – but never fixed.
- The recently re-launched Carbon Metric developed by Zero Waste Scotland uses alternative datasets to the default models included within WRATE. For example, the benefits of recycling card and the treatment of food waste in an AD facility are both handled differently. I understand that Defra has used emissions factors from the Carbon Metric, rather than those from WRATE, in its local authority dataset Waste Data Flow, suggesting it too has lost faith in the tool.
Generating errors
A glaring error – which can badly mislead those who rely on the tool – lies in the default assumption regarding the carbon intensity of the UK’s current and future electricity generation. When evaluating the GHG impacts of new facilities that generate electricity, like energy from waste (EfW) plants, it’s important to know how much CO2 would have been emitted by the source of electricity that’s being displaced: the ‘marginal source’.
DECC publishes clear guidance annually regarding the approach to take when policy evaluation requires an estimate of the GHG emissions (or ‘carbon intensity’) associated with the marginal source of electricity generation. The guidance for 2013 gives a carbon intensity similar to that of gas combined cycle gas turbine (CCGT) generation. In effect, what DECC are correctly saying is that when new generating capacity comes on stream, currently it will mainly be gas that is displaced. However, DECC shows CCGT decreasing significantly in the future, reflecting greater use of renewables to achieve our legally binding climate change targets.
Although some changes were made in WRATE V2 to the electricity grid mix, the default values in WRATE still assume for 2015 a marginal generation mix of 47% coal (far less efficient) with remainder being CCGT, whilst the mix for 2020 assumes 34% coal.
It isn’t easy to establish the original source of these values. I obtained an apparently unpublished report by the EA’s technical advisors ERM dating from 2005, which suggests they were developed by ERM in 2004 using a dataset produced by the then Department of Trade and Industry, making them a decade old.
These ancient defaults have major implications if you’re modelling the future performance of EfW facilities. Future performance is a significant consideration, as a facility procured today would be expected to operate well beyond 2020.
WRATE’s defaults regarding the marginal source of electricity generation overstate the emissions savings of an incinerator operating in 2020 by around 0.42 kg CO2 per kWh of electricity generated compared with the approach recommended by DECC. For 2020, WRATE therefore overestimates the annual benefit of an EfW facility treating 250,000 tonnes of waste at an efficiency of 25% by around 50,000 tonnes of CO2 equivalents. Beyond 2020, the discrepancy just gets worse.
Storing up trouble
EfW’s apparent benefits are also boosted by WRATE’s out of date assessment of how biogenic carbon (carbon of plant and animal origin) behaves in landfill. Reasonably enough WRATE follows IPCC guidance on modelling landfill impacts, which suggests over half of this carbon doesn’t appreciably degrade in 100 years. The European Commission’s Handbook on LCA now says this ‘temporary’ storage of biogenic CO2 should be accounted for within the results through the application of a credit reflecting the amount of carbon stored, but WRATE doesn’t do this – it ignores all biogenic carbon. Recently published research confirmed the need for a further adjustment because biogenic methane has less impact than that originating from fossil sources – again, not done in WRATE.
Let’s assume that residual waste includes approximately 20% paper and 25% food waste. Compared with the latest methods, WRATE’s analysis of the effects of sending 250,000 tonnes of waste to landfill in a particular year now overstates the emissions by around 70,000 tonnes CO2 equivalent over the following hundred years.
Using energy wisely
Part of the problem with keeping WRATE updated would appear to be that no government department currently ‘owns’ the tool. While the EA is “exploring the options for the future management of WRATE”, recent substantial cuts in Defra and EA budgets don’t bode well. The outdated energy mix data could be remedied relatively easily via a regular email update to the WRATE user community. However, many of the other issues I’ve explained are not easily fixed – remodelling the landfill impacts or rebuilding the background databases would require costly work. And even expert modellers can’t access the landfill model to make adjustments.
In the current climate of ever reducing public expenditure, it seems to me WRATE’s expensive front end with its snappy little pictures of waste trucks are a luxury we can ill afford. We should scrap WRATE and replace it with a much simpler – but properly maintained – spreadsheet tool. It should focus on the climate change impacts of waste management, which is the main thing WRATE is used for, instead of pretending to be comprehensive – I have in mind something similar to the US Environmental Protection Agency’s WARM model. As things stand, WRATE does more harm than good, and government bodies that rely on it to inform policy choices are being given false confidence that they are making the right decision.
Are we really getting 53.46% of our electricity from nuclear by 2035. That is not a realistic figure. 2020 figure is 4% too high. Most data is from 2006, thats getting on for 20 years old, a life time in LCA.
Hi Ann
A good contribution to the debate, I agree that one of the (low cost!) approaches that should be adopted for WRATE is to provide the latest Government position on key variables (Energy mix etc) that could be applied into the WRATE model to move its results in line with current thinking. The Green Investment Bank already do this in terms of specifying marginal energy mix which is readily altered in WRATE.
I have also long campaigned for a list of process errors to be circulated to the WRATE user community, as I am sure you (as I) have also discovered several processes within WRATE do have erroneous allocation rules etc. We report them to the EA when we have discovered them and I presume you / others do similar, but it is never circulated. This must be a valuable but low cost method of improving the model usage.
The issue I see with developing a simple spreadsheet model is that WRATE is predominantly applied (in my experience) in procurement or more recently GIB works, when its functionality allows development of complex new or bespoke technologies within its comparative framework. A simple spreadsheet would not be able to achieve this and to try and emulate that capability would cost a lot more than improving elements within WRATE in my opinion.
Having an alternative spreadsheet based system (akin to the GLA model – although this uses WRATE defaults or WARM) would be useful for waste strategies and plans which require more generic outputs, however again these seem less in demand in straightened times.
I am not sure you can assume that as Government has not used WRATE in all its carbon calculations throughout it represents a loss of faith – it may just be imperfect Government (!) – however I don’t have any Defra / EA insight on this point.
I agree with you that in challenging financial times there may be a temptation to drop WRATE and make something new or alternative, but I would see that a more preferable solution (and I would think much more cost effective) would be to bring a panel or steering group together to seek to improve the model which has very strong functionality but I agree would benefit from a refresh of key datasets. I would think that Eunomia with your in-house LCA model plus WRATE experience would be excellent contributors to such a forum, as would other WRATE users / academics / LCA practitioners etc.
The costs of developing an alternative functioning LCA model for waste management is not only in its development but also in training of users, there has been considerable investment in training WRATE users to Standard or Expert level over the past six or so years, so I wouldn’t underestimate this aspect either, when considering if a different model should be developed.
This could be an area that one of the umbrella organisations could take forward (e.g. CIWM / ESA) or perhaps with a University / EU funding?
Thank you for raising the topic, it is an important area and I believe WRATE needs on-going updates and maintenance (including ownership!).
Ann
Really interesting article. I have an interest here in the inevitable yet logical conclusion of putting incomplete data in, and getting poor judgements out – specifically when systems, and economics, are complex and regionally specific.
For balance, in your view, does WRATE also overestimate the contribution of recycling – both open and closed loop – in terms of GHG emission savings?
Simon
Robust – and recent data – on the benefits of recycling remains relatively difficult to obtain for some materials. The literature suggests that the data in WRATE is probably reasonable for metals, dense plastics and textiles. Some relatively recent sources indicate WRATE may overestimate the benefits of recycling paper, and underestimate the benefits of recycling glass. However, a more substantial problem in my opinion is that the standard models in WRATE significantly overestimate the impacts of AD (through underestimating the amount of electricity generation and the potential benefits of digestate).