Rainwater harvesting systems for toilet flushing – do they ever make environmental sense?

Why exactly we SHOULD be flushing drinking water down the toilet.

Whilst working at the Centre for Alternative Technology (CAT), near Machynlleth, I managed a number of rainwater harvesting systems, in which rainwater was collected from the roofs of buildings, stored in tanks and then pumped round the building to flush toilets with. At first sight, these systems make a good deal of sense; flushing drinking quality water down the drain would be daft, right? Unfortunately not. Let’s examine a few of the key arguments.

“It’s a wasteful of energy and resources to treat water to potable standards only to flush it down the drain”

Water treatment does indeed require energy and resources. However, in the UK at least, it’s a very efficient process compared to other parts of the water use cycle. The diagram below is a pie chart showing the carbon impacts of the entire cycle of water abstraction, treatment, distribution, collection of sewage and then treatment of that sewage before it is released into the environment.

EA carbon emissions from water use

Environment Agency (2008). Greenhouse gas emissions of water supply and demand management options. Science Report SC070010

This is a very surprising picture to most people. The GHG emissions associated with taking water from the environment, cleaning it, storing it and moving it to our houses is 4% of the total; insignificant compared to the GHG emissions arising from our use of it in the home (89%). So if we are installing a RWH system we addressing a very small part of the overall problem; if we were interested in greenhouse gas emissions relating to water use, we would be focussing on the 89% of emissions that relate to water use in the home. This is largely from our use of hot water, and I deal with this in a separate blog post here. And for those that believe that ‘every little helps’, as we will see below, even if we are interested in this very small part of the pie chart, it is fairly easy to demonstrate that RWH systems have higher energy costs than mains water.

“But it’s about saving water, not energy”

In order for this argument to make any sense, we need to consider where the rainwater is being saved, and whether we live in a water scarce area that could benefit from us using less mains water. And if we are looking for a supplementary water source, we would do well to start with one that is plentiful. Yet with rainwater there is a fundamental mismatch between supply and demand. The areas of the UK with high rainfall (where a RWH system would reasonably be expected to provide useful amounts of water) are not those that suffer from water scarcity (see the maps below). In addition to the basic geographical problem, there is also a seasonal mismatch between supply and demand. None of this should be surprising; our mains water supply in the UK is dependent on rainfall, either stored in reservoirs, or via replenishing groundwater supplies.

water maps

On the left is a water scarcity map (Environment Agency), and on the right is a map of rainfall (Met Office). On the water scarcity map, red and amber areas are those that are water scarce. On the rainfall map, areas that are dark brown are those with low rainfall.

“They help protect against flooding”

This is another argument often used by proponents of RWH, but again, it doesn’t really stand up to much scrutiny. There are several different types of flooding, and since each has a different cause, we need to understand the cause before assuming that a particular measure will mitigate against flooding. I have written about this here. The type of flooding which could potentially be mitigated by RWH is pluvial; these flooding events are typified by short duration, high intensity rainfall in areas with a high proportion of impermeable surfaces, and result in very sudden overload of surface water drainage systems. Temporary storage of rainwater in attenuation structures, which can then be released after the peak event has passed is therefore a frequently specified part of a mitigation system. Most often, these attenuation structures are large basins, or sometimes swales (wide shallow ditches that both hold a lot of water and are constructed to allow it to flow to an area where it would cause less damage). In some instances however, these attenuation structures are tanks. Proponents of RWH systems often claim that because the RWH tank stores rainwater, the RWH system contributes to flood prevention. However, this completely neglects the scale of flood attenuation systems compared to RWH systems; flood attenuation basins are designed to accommodate hundreds of cubic metres, in contrast to a domestic RWH system which might have a tank of 1-2 cubic metres. Additionally, in order to provide water for toilet flushing, the RWH system tank needs to contain water, but in order to protect against flooding, the tank needs to be empty and ready to recieve water. So we have both a scale problem, and a timing problem. It is possible to imagine a RWH system that was programmed to empty its tank in response to a local weather forecast (i.e. immediately prior to an expected pluvial flooding event), but this does not address the issue of scale.

“They will save you money”

RWH system brochures will typically quote a price for a system, and then using an assumption about how much mains water you will save, calculate a simple economic payback. This misses out some fairly obvious expenses, such as servicing costs, and the replacement of components when they fail. Assumptions about how much water will be saved and what the price of mains water will be in the future are also required. If you are considering a RWH system, I would strongly recommend you did your own maths; even the cost of an annual service is likely to be more than the cost of the water that you would save. An entire PhD thesis was devoted to analysing the economics of RWH systems in detail. Richard Roebuck at the University of Bradford created an economic model of the whole life costs of RWH compared to mains water and analysed several thousand potential scenarios. The results are shown below.


On the x axis we have the whole life cycle costs of a domestic water system that includes RWH. On the y axis is the whole life cycle costs of using only mains water. Each of the triangles represents a different scenario (volume saved, maintenance frequency etc). Triangles to the top left of the diagonal line are scenarios in which RWH worked out cheaper than mains water. As you can see, there aren’t any! The journal paper describing Richard’s work is available here. Whilst the actual costs will have changed since this peice of work, the basic message remains the same.

Does it work at scale?

In the UK? Possibly, although nobody has yet produced data of an installed system that makes environmental sense. Some years ago when I undertook a life cycle assessment (LCA) on a proposed installation (a school with a roof collection area of 3,100m2, and a UK average rainfall of 1 metre), the greenhouse gas emissions from the scenario with a RWH system were around 7 times higher than in the same scenario using just mains water. This was partly due to resources required to install the capital infrastructure (tanks, pumps, additional pipework) and also due to the electricity required to run the pumps in the system. A copy of this study is available here. A review of a number of similar studies on RWH systems is available here.

Obviously, the answer to questions of scale depends on our point of comparison. Intercepting water and distributing it is exactly what our mains water system does; reservoirs capture surface water in areas of high rainfall, it is then treated and distributed to households. The fact that domestic scale RWH makes no sense in the UK is as much to do with the excellent mains water infrastructure we have as it is to do with poorly designed rainwater systems. But it is possible to imagine a situation where a farmer was managing a large land area, and entered into discussions with a water company regarding how best to manage the runoff of local rainfall. We are at risk of taking an ideological view of small scale compared to large scale, but for most systems, be they water, electricity, education, health, we would do well to think about the appropriateness of scale.

Is RWH a good idea anywhere?

Yes! As stated above, the critical considerations are what our alternatives are, and what an appropriate scale might be. Town scale RWH serving all domestic water uses could be an excellent solution in countries where the population is too dispersed for mains water infrastructure. There are many countries in which a household scale RWH system is the most practical way of supplying water. In countries that are reliant on desalination for public water supplies, the energy intensiveness of water treatment might also make the use of a lower quality water supply appropriate for toilet flushing, although other water sources could also be used if dual plumbing systems were installed (e.g. as in Hong Kong, where dual plumbing systems exist, one of which carries non-potable water, in this case sea water).

In the UK we have an excellent mains water supply. If you are interested in the environmental impacts of your water use you would do better to remember the hierarchy of reduce, reuse, recycle. In the case of water in the UK, water efficiency measures are to be recommended to reduce your mains water use, but taking steps beyond this are of very limited environmental benefit.

A chapter of my book Choosing Ecological Water Supply and Treatment discusses RWH in more detail. There are also chapters on grey water systems and garden water use.



2 thoughts on “Rainwater harvesting systems for toilet flushing – do they ever make environmental sense?”

  1. Nice article with some good references to further reading. If I was being picky I’d question how the first diagram (the pie chart) is relevant to the “toilet flushing” in the title since with the possible exception of clothes washing, none of the permitted uses of harvested rainwater involve it being heated. I take the point that heating water in a more environmentally friendly manner (or using less of it in the first place) is a much better first step to take than switching to harvested rainwater, but some readers will already have addressed that, e.g. by using excess solar PV generation to heat DHW.


    1. Glad you found it interesting. Perhaps I didn’t explain why the first pie chart was there particularly well; if we’re trying to address the impacts associated with water use, then starting with the biggest part of the pie is easiest. When we look at the impact of water supply and treatment and see it’s a small part of the pie, it should serve as a warning that any savings we make are marginal in the overall scheme of things.

      I’d question whether anyone produces enough excess PV to make heating DHW anything approaching carbon neutral, and I guess my preference would be to export excess electricity to the grid wherever possible. Or to install SWH rather than PV (I’ve not read anything simple on how to make this decision though, so if you find something, do let me know).


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: