ANALYSIS OF HOUSEHOLD WATER-USE BEHAVIOR FOR USE AS
IAQ MODEL PARAMETERS
CR. Wilkes1*, SC Hem2**, JN Blancato2, AD Mason1, and LL Niang1
1 Wilkes Technologies, Inc, Bethesda, MD 20814
2 US Environmental Protection Agency, Human Exposure and Atmospheric Sciences Division
ABSTRACT
In assessing exposure to indoor air contaminants, the understanding of population water-use
behavior for indoor water-use activities as a function of demographic characteristics is vital to
obtaining realistic exposure estimates. In response to the need for more specific and usable
population based water-use data for exposure modeling, new data sources have been
analyzed. Frequencies, durations, volumes and flowrates of showers, baths, clothes washers,
dishwashers, toilets and faucets are presented, derived from analyses of the National Human
Activities Pattern Survey (NHAPS) database, the Residential End Uses of Water Study
(REUWS) database, the Residential Energy Consumption Survey (RECS) as well as from
current literature and manufacturer information. Distribution parameters for water use
behaviors are identified for use as inputs for modeling exposure to water borne contaminants.
This work has been funded wholly or in part by the United States Environmental Protection
Agency and has been approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
INDEX TERMS
Water use, Activity patterns, Modeling, Exposure, Indoor air
INTRODUCTION
A realistic assessment of exposure and risk to water-borne contaminants requires accurate
summaries of water usage patterns. Currently, there are a few limited studies that analyze
household water-use behavior; the leading resources are summarized in the Exposure Factors
Handbook (U.S. EPA, August 1997). In response to the need for more specific and usable
population based water-use data to be used as inputs for exposure modeling, water-use data
from various databases and literature sources were extensively analyzed in the recent USEPA
report entitled "Quantification of Exposure Related Water Uses for Various US
Subpopulations" (Wilkes, 2002) to quantify water-use behavior for a variety of demographic
groups for use in assessing exposure to water-borne contaminants. This paper summarizes
those findings on the frequency, duration, volume and flowrate characteristics for showers,
baths, clothes washers, dishwashers, toilets, and faucets.
Data sources analyzed for water- use behavior parameters are: The National Human Activity
Pattern Survey (NHAPS) compiled from a 1992-1994 telephone 24-hour recall survey of a
representative sample of the US population (over 9,300 respondents) (Klepeis et.al., 1996);
the Residential End Uses of Water Study (REUWS) compiled from a 1996-1998 study of
water-use data recorded by a data-logging device attached to the household water meters of
Contact authors* emails: c.wilkes@wilkestech.com and Hem.Stephen@epamail.epa.gov
-------
approximately 1200 single-family homes (100 homes in 12 different municipalities) (Mayer
et.al., 1998); and the Residential Energy Consumption Survey (RECS) compiled from a 1997
nationwide survey focusing on household energy use information (5,900 respondents)
(USDOE, 1995). In RECS, household data were weighted to statistically represent the 101.5
million U.S. housing units. Statistics presented herein are calculated using the weighted data.
DISCUSSION
When applicable, the frequencies and duration data from NHAPS, REUWS and RECS, sorted
by demographic characteristics (such as education, employment status, U.S. EPA region, etc),
were analyzed and compared for each type of water use. After comparing the databases, it
was concluded that databases based on recall surveys, like NHAPS and RECS, are reliable
sources for frequency information of relatively infrequent events such as showers, baths,
dishwasher and clothes washer use, but are unreliable in reflecting more frequent events such
as faucet use. REUWS, which is based on analysis of waterflow signatures of household
water meters, is an excellent source for water-use duration information. Overall, due to the
manner in which the databases were compiled, NHAPS data are more reliable than REUWS
for frequency information, while REUWS data are more reliable than NHAPS for duration
information. NHAPS was compiled from a recall telephone survey of the respondents*
activities of the previous 24 hours. Respondents appeared to be able to remember how many
showers and baths they took, while they had difficulty estimating the durations of the events,
as the duration values appeared to be overestimated and clustered around 5-minute intervals.
Furthermore, the data was truncated as all survey responses for shower duration over one hour
were recorded as "greater than one hour" instead of the actual number of minutes. In contrast,
REUWS was compiled from direct mechanical measurements of water usage logged at
household water meters and subsequent waterflow disaggregation by the software program,
Trace Wizard, to determine individual water uses. REUWS contains measured values of
duration, volume, and flowrates of the water-use events in its database. For this reason,
REUWS provides very accurate duration data. However, REUWS has a few integral
limitations that make it less reliable in reference to frequency data, such as the inability to
discern which person is performing the water uses in question, and at times Trace Wizard
mislabeled events, reporting water uses that were clearly unrealistic. In regard to the
frequency of clothes washer and dishwasher use, the RECS database was the most reliable
source as the survey questions were more straightforward than those asked for NHAPS.
Dishwasher and clothes washer durations and volumes are best characterized using a
combination of data from REUWS, manufacturers' data, and data from field experiments.
Only REUWS provides usable information on faucet and toilet use.
SHOWERS AND BATHS
Shower and Bath Frequency: Frequency statistics resulting from NHAPS analysis on shower
and bath use are believed to appropriately represent the populations' behavior. Table 1
presents these results for men, women, and children ages 5-12 yrs. Although the impact is
believed to be relatively small, potential biases must be recognized including the ability to
recall events and biases due to perceived societal expectations. The overall household bath
frequency based on the REUWS analysis is also presented for comparison purposes.
Shower Duration: Shower duration statistics resulting from REUWS analysis are believed to
appropriately represent the length of showers for the given population. The lognormal
distribution provides a good representation for the shower duration data since the data are
positively constrained. The geometric mean and standard deviation, and arithmetic mean are
presented in Table 2 for the various demographic groups. The shower durations reported in
-------
NHAPS were found to be unreliable, as in comparison to other sources they appeared to be
overestimated and 89% were reported at a five-minute interval (i.e., 5, 10, 15, or 20 minutes).
Bath Duration: NHAPS contains the best available dataset for bath durations, since surveys
like REUWS contain only the amount of water used to fill the bathtub not the bath duration.
Although there are significant biases in the dataset, the duration statistics are recommended
until a more definitive study provides better information. The durations reported in NHAPS
are biased by a multitude of factors, including round-off errors as approximately 94% of baths
were reported with durations at a five-minute interval, and estimation errors whereby a
comparison with other shower duration studies indicates that duration responses in NHAPS
tended to be overestimated. The NHAPS bath duration data are fitted to a lognormal
distribution with results shown in Table 2.
Shower and Bath Volume and Flowrate: REUWS shower volume and flowrate data were
analyzed and fit to lognormal distributions as shown in Table 2. However, as with other
REUWS data, these data may be impacted by misclassification and single events reported as
multiple events. Bath fill behavior and volume is not well enough understood to make
recommendations based on the REUWS data. However, general dimensions of standard
bathtubs are well understood, holding approx. 210-250 Liters (55-65 gallons) of water, when
filled to the overflow, which is likely reduced by 20-30% due to the bather's body volume.
In general, the analysis showed that age significantly influenced shower and bath frequency
and duration. The frequency of showering and bathing reported in NHAPS agreed reasonably
well with previous studies; however, durations of these events were found to be significantly
longer. The frequency of showering reported in REUWS was slightly less than that reported
for NHAPS, though this may be due to NHAPS reflecting all showers taken during the day
including those taken at work or at health clubs, while REUWS only recorded showers taken
at home. The durations of showers reported in REUWS are consistent with other studies.
Table 1. Mean Frequency of Shower and Bath Use (Events per Person per Day)
Data Source
Shower Frequency
Bath Frequency
Men(N)
18-48yrs
NHAPS
1.24(1023)
0.21 (1021)
Women (N)
18-48yrs
NHAPS
1.12(1086)
0.38 (1082)
Children (N)
5-12yrs
NHAPS
0.55 (329)
0.48 (336)
A11(N)
NHAPS
0.98 (4608)
0.32(4591)
A11(N)
REUWS
0.82 (2947)
Unknown
CLOTHES WASHERS
Frequency, The RECS database proved to be the most reliable resource for clothes washer
use frequency data, as its data directly reflects the number of loads of clothes washed in the
household per week. In contrast, the NHAPS data was not useful for two major reasons: the
data reflected only the washing done by the survey respondent, and it was not clear whether
the answer reflected the number of loads washed or the number of days per week the wash
was done regardless of the number of loads done each day. Based on an analysis of RECS
data, the number of loads of laundry washed per household per week increases with an
increasing number of occupants in the household as shown in Table 3.
-------
Table 2. Duration, Volume and Flowrate Statistics for Showers and Baths
N
Geometric
Mean
St. Dev.
Arithmetic
Mean
Showers (All, REUWS)
Shower Duration (min)
Shower Volume, L (gal)
Shower Flowrate, L/min (gal/min)
3241
3241
3241
6.8
59.8(15.8)
7.6 (2)
0.49
0.56
0.46
7.7
73.1 (19.3)
9.1 (2.4)
Baths (NHAPS)
Bath Duration : All (min)
Men, 1 8^8 yrs (min)
Women, 1 8-48 yrs (min)
Children, 5-12 (min)
Bath Volume, L, (gal)
Bath Flowrate (REUWS) L/min (gal/min)
4591
1020
1081
328
n/a
1966
17.6
17.2
17.8
18.6
n/a
16.6(4.4)
0.63
0.69
0.72
0.51
n/a
0.54
20.9
20.8
21.5
20.8
189(50)*
18.5(4.9)
*estimated based on typical bathtub size
Table 3. Mean Frequency of Clothes Washer Use (RECS, N=5,900 households)
Number of Occupants
Per Household (loads/week)
Per Capita (loads/week)
1
3.2
3.2
2
5.2
2.6
3
6.8
2.3
4
8.5
2.1
5 or more
9.2
1.8
Overall
6.1
2.3
Clothes Washer Duration and Volume. In regard to duration, REUWS provides data on the
durations of the individual cycles (washes and rinses), which can be combined to determine
the time it takes from the start of the first fill until the end of the last fill. However, REUWS
does not provide data on the duration of the entire event, which would include the time to
complete the final agitation and spin. In order to characterize the entire clothes washer
duration, various sources were analyzed. For individual cycle duration information (wash fill,
rinse fill), the REUWS data was used. For information on the agitation and spin durations,
data from timed experiments on two typical top-loading machines were used. These values
are provided in Table 4. According to the REUWS data, the fill (1st cycle) and first rinse (2nd
cycle) are 100% likely to occur. The second (3rd cycle) and third rinses (4th cycle) are 18.7%
and 0.8% likely to occur. Weighting the duration values for these additional rinses, the total
duration of the washing event in this configuration was calculated. Table 5 presents the
average volume of water used and total duration of a typical clothes washer event based on
information presented in Consumer Reports (July 1998, July 1999, August 2000).
Table 4. Mean Volumes and Durations for Typical Top-Loaded Clothes Washer Cycles
Cycle 1
Volume
L(gal)
62.8(16.6)
Duration (minutes)*
Fill
3.8
Agitate
12.0
Drain
4.0
Cycles 2, 3, 4
Volume
L (gal)
57.9(15.3)
Duration (minutes)*
Fill
7.5
Agitate
4.0
Drain
8.0
Total**
Duration
43.1 min
* Based on experiments on Kenmore 70 series 1992 and 90 series 1999 set to high-water level.
** Cycles 1 & 2 are 100% likely, cycle 3 is 18.7% likely, and cycle 4 is 0.8% likely (REUWS).
-------
Table 5. Mean Volumes and Durations for Clothes Washer Events
Machine Type
Top-Loading (before 1999)
Top-Loading (2000 models)
Front-Loading
Volume
L/load (gal/load)
155(41)
124.9(33)
102.2(27)
Duration
min/load
43
45
64
Citation
Consumer Reports (July 1998, July 1999)
Consumer Reports (Aug. 2000)
Consumer Reports (July 1998, Aug 2000)
DISHWASHERS
As compared to other water sources in a household, dishwasher uses represent a relatively
small source in respect to water-bome contaminant exposure because of the infrequent usage,
small water volume, and the relatively sealed washing compartments. As such, the exposure
resulting from dishwasher use can be expected to be a very small portion of an occupant's
overall exposure to water borne contaminants.
Frequency of Dishwasher Use. To represent the frequency of dishwasher use, the most
reliable data was judged to be from the RECS analysis, because the RECS survey question
reflected household use, while the NHAPS survey question reflected dishwasher use of only
the respondent. However, the RECS data did not capture the lower frequencies of use, as the
response choices were either: "Daily", "4 to 6 loads per week", or "less than 4 loads per
week." In analyzing this data, the frequency was calculated assuming the median value for
each frequency range. Considering that 56.3% of the respondents answered "less than 4 loads
per week", this data clearly lacks definition. According to the RECS analysis, shown in Table
6, frequency of dishwasher use per capita decreases as family (or household) size increases.
Table 6. Mean Frequency of Dishwasher Use (RECS, N=5,900 households)
Number of Occupants
Per Household (loads/week)
Per Capita (loads/week)
1
2.5
2.5
2
3.4
1.7
3
3.8
1.3
4
4.6
1.2
5 or more
5.1
1.0
Overall
3.7
1.4
Dishwasher Duration and Volume. Based on the information available from dishwasher
manufacturers, the typical dishwasher event is comprised of approximately 5 fills (incl.
washes and rinses). Assuming the "Normal Wash" option across brands listed in various
manufacturer-supplied data and from Consumer Reports (March, 1998), the overall average
duration of a dishwasher load (including drying time) was 100 minutes and the total volume
of water used was 30.3 liters (8 gallons). See Wilkes, 2002 for details.
TOILETS
Analysis of REUWS data provides reliable information for toilet flush frequency, and toilet
tank fill duration, volume and flowrate. The NHAPS survey did not include questions about
toilet use. From the REUWS data analysis, it is estimated that, on average, a person flushes
5.5 times per day. However, it must be kept in mind that the REUWS database only contains
water-uses that occurred in the home and therefore doesn't account for toilet uses at work or
elsewhere during the day. The amount of water that toilets use to flush has dramatically
decreased due to conservation efforts and mandated plumbing codes. Early models used about
19-26 liters (5-7 gallons) per flush, while newer toilets manufactured after 1992 are required
by U.S. law to use only 6 liters (1.6 gallons) per flush. The summary results from the REUWS
analysis on toilet tank fill volume, duration, and flowrate are presented in Table 7. It is safe to
assume that as years pass, the average volume of water used per flush in any given U.S.
population will decrease as older toilets are replaced with more efficient toilets.
-------
Table 7. Toilet Flush Characteristics (REUWS)
Parameter (N)
Frequency (22,582 days)
Toilet tank volume (245,382 flushes)
Toilet tank fill flowrate (245,382 flushes)
Toilet tank fill duration (245,379 flushes)
Mean
5.5 flushes/person/day
13, 2 liters (3 .48 gallons)
14.8 liters/minute (3.9 gallons/minute)
71.4 seconds
St. Dev.
3.23
1.18
1.31
29.77
FAUCETS
Faucet usage is probably the most difficult household water use to characterize in general
terms because each water use may differ greatly from the next in its duration, volume,
flowrate and temperature. However, the REUWS database is the best available source of
frequency, volume, duration, and flowrate information regarding faucet use. Analysis shows
that frequency of faucet use is dependent on the number of occupants in the household, as the
mean faucet uses per person per day decreases as the household size increases (Wilkes, 2002).
This results from the many faucet uses that are house-related not individual-related, such as
for cooking or cleaning. The general characteristics of faucet use are presented in Table 8.
Table 8. Faucet Use Characteristics (REUWS)
Parameter
Frequency, uses/person/day
Volume per event, L (gal)
Duration per event, seconds
Flowrate, L/min (gal/min)
N
965 people
973,722 uses
973,71 7 uses
973,722 uses
Mean
17.4
2.6(0.7)
33.9
4.5(1.2)
Mode
—
0.38(0.1)
10.0
1.9(0.5)
Min.
—
0.0
10.0
0.0
Max.
-
142.3(37.6)
5400.0
40.5(10.7)
St. Dev.
11.6
3.8(1.0)
45.6
2.6(0.7)
CONCLUSIONS
Linking the use of contaminated water with exposure and potential risk analyses can be
accomplished using an exposure model that represents factors leading to contaminant release
and contact. Such a model must represent the physical environment, the emission
characteristics of the water appliances during their use, and the water-use and location
behavior of the occupants. The water-use characteristics and distributions discussed and
presented in this paper (from Wilkes, 2002) are analyzed such that the data can effectively be
utilized by an exposure model (such as the Total Exposure Model (TEM)) when simulating
realistic occupant water-use behaviors of various populations. Though the available databases
provide significant information on water-use parameters, there is need for considerably more
research in quantifying these behaviors.
REFERENCES
Consumer Reports. Consumers Union of U.S. Inc. Yonkers, NY. August 2000. Vol. 65. No. 8;
July 1999. Vol. 64. No. 7.; Mar. 1998. Vol. 63. No. 3.; July 1998. Vol. 63. No. 7.
Klepeis NE, Tsang AM, and Behar JV. July 1996. Analysis of the National Human Activity
Pattern Survey (NHAPS) Respondents from a Standpoint of Exposure Assessment^
EPA/600/R-96/074. U.S. EPA, National Exposure Research Laboratory, ORD.
Mayer PW, DeOreo WB, Opitz EM, et al. Residential End Uses of Water. AWWARF Report.
USDOE. 1995. Residential Energy Consumption Survey (RECS): Housing Characteristics 1993.
DOE/EIA-0314(93), U.S. DOE, Energy Information Administration, Wash. DC.
USEPA. Aug. 1997. Exposure Factors Handbook Vol. III. EPA/600/P-95/002Fc.
Wilkes CR, Mason AD, Niang LL, et al. 2002. Quantification of Exposure Related Water
Uses for Various U.S. Subpopulations. EPA Draft Report.
-------
NERL-RTP-HEASD-02-071
TECHNICAL REPORT DATA
1. Report No.
EPA/600/A-02/072
4. Title and Subtitle
Analysis of Household Water-Use Behavior for Use as IAQ Model Parameters
5. Report Date
Submitted 4/02
6. Performing Organization Code
7. Author(s)
Charles R. Wilkes1, Stephen C. Hem2, and Andrea D. Mason1
8. Performing Organization
Report No.
9. Performing Organization Name and Address
1. Wilkes Technologies, Inc, Bethesda, MD 20814
2. U.S. EPA, Las Vegas, NV 89114
10. Program Element No.
3906, 8.2.1, 3-002 & 3-025A, APG28,
APM36
11. Contract/Grant No.
IAG #DW4793944301
12. Sponsoring Agency Name and Address
U.S. EPA, Las Vegas, NV 89114
13. Type of Report and Period
Covered
14. Sponsoring Agency Code
15. Supplementary Notes
To be presented at Indoor Air 2002, Monterey, CA, June 30 - July 5, 2002
-------
TECHNICAL REPORT DATA
16. Abstract
Chemicals brought into the home through the domestic water supply, result in human exposure via the three principal
routes: ingestion, inhalation, and dermal contact. Disinfection byproducts resulting from the treatment of municipal water
supplies lead to the formation of a mixture of organic compounds ranging from high to low volatility. The inhalation route
has been shown to be the most significant route of exposure for the more volatile chemicals.
In assessing exposure to indoor air contaminants, understanding population water-use behavior for indoor water-use
activities as a function of demographic characteristics is vital to realistic exposure estimates. Currently, there are few and
limited studies that analyze household water-use behavior; the leading resources are summarized in the Exposure Factors
Handbook (U.S. EPA, August 1997). In response to the need for more specific and usable population based water-use
data to be used as inputs for exposure modeling, new data sources have been analyzed. Frequencies and durations of use
of showers, baths, clothes washers, dishwashers, toilets and faucets are presented and compared for various demographic
groups derived from analyses of the National Human Activities Pattern Survey (NHAPS) database, the Residential End
Uses of Water Study (REUWS) database, the Residential Energy Consumption Survey (RECS) as well as from current
literature and manufacturer information. Volumes and flowrates are also analyzed from REUWS for the various water
uses. Distribution parameters for water use behaviors of various sub-population groups are identified for water uses to be
used as inputs for modeling of exposure to water borne contaminants.
The results from the database analyses will be presented in the form of fitted distributions for water-use durations for the
various sub-population groups. A summary of frequency of use parameters for all water uses, and volume and flowrate
characteristics of clothes washers and dishwasher appliances are also presented.
This work has been funded wholly or in part by the United States Environmental Protection Agency and has been
approved for publication. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
17. KEY WORDS AND DOCUMENT ANALYSIS
A. Descriptors
B. Identifiers / Open Ended
Terms
C. COSATI
18. Distribution Statement
Release to the public
19. Security Class (This
Report)
Unclassified
20. Security Class (This
Page)
Unclassified
21. No. of Pages
6
22. Price
------- |