Water saving in domestic car washing
David Phipps1, Rafid AlKhaddar2 and Maxine Stiller3
1Built Environment and Sustainable Technologies Research Institute
(BEST), Liverpool John Moores University, Byrom St, Liverpool L3
3AF, UK email firstname.lastname@example.org.
2BEST, Liverpool John Moores University, Byrom St, Liverpool L3
3AF, UK email email@example.com.
3 United Utilities Group PLC, Lingley Mere Business Park,
Warrington, WA5 3LP
There is considerable interest in promoting domestic water
efficiency. A wide range of studies have focussed on water saving
within the house, both by technical means (low flush WCs, reduced
flow in showers etc.) and by changing user habits. However, rather
less attention has been paid to water use outside the house. One of
the activities which uses water outside the house is car washing.
Much is known about water use in commercial car washes but
relatively little about domestic car washing. The aim of this study
was to gather information about the water used for this activity. Key
questions were - whether the wash is done in a commercial car-
wash or at home and if the latter what is the volume used per wash
and the frequency of washing? As part of this trial the use of water
on different types of cars using different washing regimes was
measured. From the study conclusions are drawn about the likely
water use for domestic car washing.
Water efficiency is a topical and important subject. United Utilities
PLC (A Northwest England Utility Provider) has previously supported
work to examine water use in the domestic setting in an attempt to
identify water-saving measures. The work reported here continues
this work and has examined a previously under-reported area of
domestic water use, namely car-washing.
Information on personal habits on car-washing was sought using a
questionnaire which was designed in conjunction with United
Utilities. Information was gathered on a range of factors which might
influence car-washing habits. The questionnaire was distributed to
staff at LJMU (Liverpool John Moores University) and United Utilities,
by group e-mail. The questionnaire was distributed and responses
collected and analysed in accordance with data protection and ethics
policies of LJMU. Respondents were self selected.
2. Car wash trials
Car wash trials were carried out using a proprietary car shampoo
and wax combined product, following the manufacturer’s
instructions. Factors varied were type of car, method of washing and
person carrying out washing. Additionally some attention was paid to
the condition of the car.
i: Type of vehicle
A range of vehicles were used, broadly categorised into five types*,
as shown below:
1. Small car e.g . Nissan Micra
2. Medium car e.g. Vauxhall Astra
3. Large car e.g. Ford Mondeo
4. People carrier e.g. Peugeot 807
5. Large 4x4 e.g. Range Rover
*The actual cars used varied as available.
ii: Condition of vehicle
1. Lightly soiled
2. Heavily soiled
iii: Method of washing:
Three different methods of washing were used and several individual
washers. Water use was measured using a standard domestic water
meter. Typical results are shown below in Tables 1 and 2.
1. Apply carwash
from bucket with sponge. Rinse off with
buckets of water
2. Apply carwash from bucket with sponge. Rinse off with hose
3. Apply carwash from bucket with sponge. Rinse off with power
iv. Person doing washing
In some cases the same car was washed independently by several
people in order to determine if there was any significant person-to-
person difference in water usage.
Data from the questionnaire was examined to see which, if any,
factors could be used to predict car-wash habits. Respondents were
asked to indicate if they washed a car or cars and if so whether this
was done at home or at a commercial car-wash. The frequency of
car-washing was then sought. The data collected also included
responses on the number, type and age of the respondent’s cars,
together with the frequency of car-washing. Some information on
personal characteristics such as age and gender and occupation
was also obtained. Home location, by post code, type of dwelling
was obtained. No strong correlation with car wash behaviour and
1 Applied at a standard rate assuming 10 litres per wash. This is included in all the reported
any of the other factors assessed was found. From preliminary
analysis of the questionnaire it was found that no individual factor
was a good predictor of car-washing habits of the respondents.
Overall, it was found that about 60%
of respondents washed their
vehicle “at home”. The overall reported frequency distribution for all
washing was 0% weekly, 32% 2-3 weekly, 24% monthly and 44%
rarely (say up to three times per year). This corresponds to an
overall average of 11 washes per vehicle per year.
For respondents using a home wash, the corresponding reported
frequencies were 0% weekly, 39% 2-3 weekly, 17% monthly and
44% rarely. This also corresponds to an overall average of 11
washes per vehicle per year. For respondents using a carwash, the
corresponding reported frequencies were 0% weekly, 28% 2-3
weekly, 38% monthly and 33% rarely. This again corresponds to an
overall average of 11 washes per vehicle per year. Assumptions on
overall water use are based on these self-reported frequencies.
There has been no verification of the data.
There is a significant difference between the pattern of frequency for
home and carwash (p=0.002) with a larger proportion of home
washers washing every 2-3 weeks than those using a car wash and
a smaller proportion washing monthly. However, the overall
calculated frequencies are similar.
The proportion of respondents who washed the car at home “rarely”
(44%) could be discounted for water saving.
2. Car wash trials
Tables 1 and 2 details initial washing trials and shows that there was
a considerable variation in the volume of water used with the method
of washing. For example, the minimum individual wash volume
recorded was 22 litres per wash2,
with a small power washer on a
This figure is approximate but is likely to only require small revision on further data
This includes an allowance of 10 litres for “soaping up”.
small car to a maximum a maximum of 59 litres per wash for a hose
on the same car with similar light soiling and the same person
carrying out each wash. However, the hose wash was based on
only the wash-off of foam from the commercial wash material.
Anecdotally other users report a different method, namely wash with
fresh water first then soap-up then wash off. That method would
increase the volume used proportionately. Notably, washing with
only a bucket gives more variable results and seems to depend
much more on the person doing the wash. Other wash methods
seem less susceptible to the effect of individual washers.
4 x 4
Table 1. Mean water use (litres/car) (n= 5 for each combination)
for car size and methods of washing with the same person carry
out the wash.
Table 1 shows the general level of water use. There was relatively
little effect of car size (large medium small) on water use (p=0.58),
but significant differences between methods of wash with bucket
only and bucket and hose being similar, but significantly higher than
the power wash (p=0.01)
There was a significantly higher level of water use for the people
carrier and the 4x4. The results for the people carrier are very
different from those of other vehicles and this is possibly due the
height of this vehicle so extra water was used whilst trying to wash
Table 2 shows the variation between different washers using the
same method of washing on the same size car. The person-to-
person variations within each method are within the range of
volumes used for that method so that it can be concluded that the
effect of individuals can be discounted when estimating a grand
mean for use.
Table 2. Repeat washing of the same vehicle different people
*Includes 10 litres for soaping up +Standard 10 litre bucket
3. Overall consumption estimate
In order to estimate the annual consumption of water used in
domestic car washing it is necessary to establish for each user:
1. Frequency of washing (from questionnaire)
2. Volume of water used per wash (from wash trials)
So that: Annual consumption = frequency (washes per year) x
Volume used per wash (litres)
Given that the overall range of volumes used is quite wide it follows
that the estimates of annual usage are at best reasonable guides.
However, estimates are judged to be reasonably reliable, at worst
within say +25%, certainly secure enough to obtain a generalised
estimate of use.
Overall results were in the range of 0.33 m3 water per person per car
per year, (small car washed with a small power washer, 11 washes)
to 1.24m3 water, per person per car per year (people carrier hose
wash 11 washes per year). The maximum figure could be increased
if either a continuous running hose was used or a pre-wash and post
rinse method was employed.
A search for other data with which to compare these results was
conducted. A figure of 300 litres per wash given by Waterwise UKa is
partially misleading as it is based on continuous flow from a hose at
10 litres per minute for 30 minutes. Within tests in this trial a figure of
192 litres per wash was obtained using a hose allowed to flow at 14
litres per minute continuously whilst a hose wash trial was being
conducted. However when the hose is used with a spray control this
amount decreases substantially.
Overall water use per wash can be compared with a reported figure
for USA of 50 US gallons (190 litres) per washb or 70 US gallonsc
(260 litres) or even 240 litres per wash from an Australian sourced.
A UK source gives a value of 30litres per wash for a small care which
is in good agreement with the values found so far in this trial.
Discussion of Results:
A search shows that there is considerable information available
about commercial car washes, with trade bodies set up in most
developed countries, such as The Car Wash Association in the UKf,
The International Carwash Association in USAg and The Australian
Car Wash Associationh. In contrast there is relatively little information
about car washing at home with such information as exists being
anecdotal and not easily verifiable. Such data for water usage in
domestic car washing is difficult to interpret. Values from USA are
much higher than those found in this study, typically 200 - 300 litres
per wash compared with the average values of this trial of about 30-
120 litres per wash. This perhaps reflects larger vehicle size and
perhaps a different set of attitudes to the wash itself in the USA
compared to the UK.
One exceptional value reported by Waterwise in the UK gave a
figure of 300 litres per wash. However, on inspection, it was found
that this was an estimate for a hose wash with the hose left running
for 30 minutes at 10 litres per minute. This has been criticised as
unrealistic and misleading4 and indeed, it does seem an extreme
estimate. However, it is a valuable pointer to a significant factor in
this type of car wash In this study a value of 192 litres per wash
found for an unrestrained hose, which whilst rather lower than that
estimated by Waterwise is nevertheless substantially more than the
litres per wash over all vehicles found so far in this
In the AutoEXPRESS report a figure of 30 litres per wash
for a small family hatchback washed with a hose, which correlates
well with the sort of values found in this study. They also report a
value of 110 litres per wash with a bucket. In this study the bucket
wash found to be most variable and clearly dependent on the detail
of the method used. So just throwing buckets of water over the car is
most water demanding and could easily use over 100 litres per
wash, but careful application with the use of a squeegee or sponge
etc can bring this down to 30 litres per wash. Interestingly, this has
been the subject of some discussion in eco chat forums.
One factor which needs further consideration in deriving a
recommendation for domestic car wash is the environmental impact,
The actual value taken here depends on weightings used, but this seems a reasonable
It is not reported whether an allowance was made for “soaping-up”.
other than simple water use, of the wash. There is a general view
that the discharge of car wash water into the drains is unwelcomei.
However, this is really aimed at discharge into surface water drains
rather than the main sewage systems. The extent of this problem will
largely depend on the nature of the drainage as to whether the
surface water and foul water systems are separate or whether there
is a combined sewer.
The use of water in individual domestic carwash is very variable and
depends on a wide range of factors but principally car size and
method of washing for power wash and hosepipe wash. For bucket
wash the water use depends much more on the attitude and
methodology of the individual. Water use in the initial trials ranged
from 33 litres per wash to 123 litres per wash. (The unrestricted
hose gave 192 litres per wash which is an upper limit, but hopefully,
not one which is likely to be reached regularly.)
Frequency of washing is highly variable. Interestingly, none of the
respondents reported washing the car weekly and 44% reported that
that they did it rarely. Thus about half of the respondents account for
the great majority of water used. At present it has not been possible
to determine any significant characteristic which might aid targeting
of this group.
Overall water use seems to range from 0.33 to 1.24m3 per car per
year, assuming an average wash frequency of 11 times per year
over all washes. This can be compared with about 55m3 per person
per year (150 litres per day) for all domestic use. As there are
approximately 2 million cars in Merseyside and Manchester this
gives a broad estimate of a total somewhere between 0.39 and 1.44
x106 m3 water used for car wash at home per year for that region
(assuming that 60% of the cars are washed at home).
A reasonable working figure would therefore be 1m3 per car per
At this stage only anecdotal evidence from the internet is available
with which to assess the effectiveness or otherwise of “waterless”
car wash products.
The next stage in this project is to experiment with the waterless car
washing kits and evaluate their efficiency in cleaning the cars and
also how would the general public accept such a product.
The authors would like to thank United Utilities Plc for funding the
a. Waterwise UK
washing-hose-vs.-bucket). Accessed January 2013.
b. http://www.green-ct.org/carwsh.htm. Accessed December 2012.
c. Winslaw, L.(2006) Enzin Articles
Operations&id=308021). Accessed December 2012.
average-car-wash-use.html). Accessed November 2012.
e. AutoEXPRESS UK. (http://www.autoexpress.co.uk/car-
news/35439/hosepipe-ban-doesn-t-add). Accessed November
f. The Car Wash Association.
(http://www.carwashassociation.co.uk/). Accessed November
g. The International Carwash Association in USA.
(http://www.carwash.org/home. accessed October 2012).
Accessed in October 2012.
h. The Australian Car Wash Association
( http://www.acwa.net.au). Accessed October 2012.
i. The Environment Agency, UK. (http://cdn.environment-
agency.gov.uk/pmho0307bmdx-e-e.pdf). Accessed October