|The Problem With Using a Single Flow Test for Sprinkler System Design|
Water Supply Fluctuations
The Problem With Using a Single Flow Test for Sprinkler System Design
By Joseph E. Kurry, P.E., and Mark Hopkins, P.E. | Fire Protection Engineering
with experience in the sprinkler industry has come across at least one
situation where a sprinkler system is installed or tested, then does not
perform as expected. Looking into the root cause of the substandard
performance, another water flow test is sometimes performed, and the
numbers are vastly different.
when the system is not meeting design specifications, the water supply
is worse. Final acceptance testing is a bad time to find out the water
supply does not meet system demands, but it can be just as devastating
prior to installation. The typical proposed solution is to conduct
another flow test; does that truly indicate that the system can perform
during the worst case scenario?
fire protection engineers are faced with water supply challenges when
designing sprinkler systems. Fire protection systems are required to
perform under worst case scenarios including times of above-average
domestic water demand. How does the system designer compare the test
data obtained through a flow test to possible worst case conditions? How
does the designer know if the test results even demonstrate the water
supply during normal conditions? For these reasons, the idea of having
an equation for adjustments to water flow data has become an
increasingly hot topic. Although significant research has been performed
on topics that affect water supplies, including forecast (predictive)
modeling, water demands, and hydraulic modeling, insufficient research
has been conducted with the focus on developing a single comprehensive
adjustment equation for use in designing sprinkler systems.
standards have typically identified the need to account for variations
in water supplies but have not provided guidance for adjustments to
water supplies or provided safety factors. Fire protection engineers and
sprinkler contractors have historically looked to local amendments or
policies to the fire codes for adjustment factors to account for
variations in water supply conditions in many jurisdictions.
Sometimes the adjustments are known before starting a design. Sometimes they are found during the review of drawings or during testing of the systems. The problem is that the methodologies for making these adjustments vary widely and there may be little consistency from one jurisdiction to the next. Additionally, the justification for these adjustments may not even be based on fluctuations in water supplies, but account for variations between the calculated design drawings and the actual field installation.
To understand why
water supplies have fluctuations, we need to understand the different
industry practices when designing water distribution systems.
systems that are not supplied by a dedicated gravity tank, or tank and
pump must rely on the water authority for the pressure and flow to
supply the sprinkler system. A fire pump can be used to boost the
pressure from the water distribution system, but it cannot create more
water if the water supply system cannot provide the necessary amount.
For this reason, it is important to understand the design difference
between sprinkler systems designed to National Fire Protection
Association (NFPA) standards and water distribution systems in
accordance with American Water Works Association (AWWA).
designing sprinkler systems, the designer has NFPA codes and standards
regulating the minimum performance of the system. Design discharge
densities over sprinkler coverage areas are clearly identified through
requirements in the reference standards. Hydraulic calculations are
preformed from the hydraulically most remote area to the connection to
the water supply. Minimum water supply demand requirements for the
system are identified. Minimum water demand requirements for the system
do not vary over time unless the hazard being protected changes or the
system is modified. In such cases, a new analysis of the system needs to
authorities are designing water distribution systems, they do not have
AWWA codes or standards regulating minimum system pressures at specific
flows. Local regulations may require certain minimum system performance,
but these regulations are specific only for their jurisdictions and are
not universally adopted. Neighboring jurisdictions that have different
water authorities may have vastly different regulations.
the consumers of water and the usage of water vary between different
communities. This variation in the usage of water creates variable
demands that are difficult to quantify. Because the demands are
variable, water authorities typically have operational ranges within
which the systems are maintained. Through the use of elevated storage
tanks, operation of additional pumps, pressure regulating valves, and
other system components, the water authorities try to maintain their
system pressures while the demand varies. The methods these water
authorities use to maintain pressure also varies. One jurisdiction may
have manual pumps while the neighboring jurisdiction uses variable speed
drive pumps or a gravity supply system.
If the manual used to
design the water distribution system indicates that the water supply
fluctuates, why are sprinkler designers only using one set of data to
design the system? An example of this variation in water supplies is
discussed next from an evaluation of a campus water supply.
CAMPUS WATER SUPPLY STUDY
in water supplies can sometimes reveal bigger issues. The authors
completed the evaluation of a campus water supply to address what the
client thought was an aging water distribution system at its facility.
In this case, water was supplied to the campus through two separate
connections to a privately operated water distribution system. Each
water supply connection was equipped with a water meter and backflow
The evaluation was
prompted as the result of an identified water supply inadequacy found
for a single building where two tests had been conducted with differing
results. The first test, conducted by a design engineer, demonstrated
that the water supply was sufficient to meet sprinkler system demand.
The second test was conducted by the sprinkler contractor more than a
year later as part of the sprinkler system design. This test identified
an inadequacy in the water supply and raised questions because the water
supply at this location had been considered sufficient for designing
sprinkler systems prior to that test.
objective of the evaluation was to identify areas of improvement at the
campus. Hydrant flow testing was performed to facilitate evaluation of
the adequacy of the existing underground fire main piping to supply the
required flow and pressures for manual and automatic fire fighting needs
and to develop an effective friction coefficient for the underground
Testing the system
demonstrated worse performance than originally anticipated. Comparison
of the water supply results to fire protection system demands throughout
the campus indicated that the water supply was inadequate for a number
of the buildings. Static pressure readings measured throughout the
facility were lower than values historically recorded. Additionally,
residual pressure readings under fixed flow conditions (measured along
isolated distribution system paths) yielded abnormal results. The data
was used to compare measured results to similar conditions with known
Hazen-Williams coefficient of roughness values. However, the results of
the evaluations did not allow for isolated repairs or replacements on
campus as hoped. As a result, a second series of tests was conducted
when the water supply was believed to have the lowest usage and its best
performance. That was not the case; the results of this testing
revealed worse performance than previously observed.
water authority was contacted. It was determined that the water
distribution system was controlled through observation of water pressure
by an operator at the supply pumping station with a normal variation of
approximately 5 psi. Supply pressures of 45 to 50 psi were normally
maintained at the pumping station. However, the differences in the
measured static pressures at the campus during the two tests were
determined to be approximately 7 psi. It was determined that the water
authority reduced the system operating pressure over a number years and
system demands increased in the surrounding communities due to the
construction of several new facilities. As a result, the available water
supply (pressure and flow) at the entrance to the campus deteriorated
technical committees are trying to quantify daily, seasonal,
geographical, and other fluctuations to water supplies that are
encountered during water flow tests to determine the typical pressure
and flow during peak demands when tests are performed during non-peak
times. Limited attempts have been made to provide an adjustment
factor(s), which have been proposed to several NFPA technical
committees, but could not be accepted due to insufficient technical
support for these adjustments.
Fire Protection Research Foundation (FPRF) literature review, titled
"Quantification of Water Flow Data Adjustments for Sprinkler System
Design,” found several things. First, daily, seasonal, geographical, and
other fluctuations are real and can impact the data collected during
water flow tests; secondly, there is relatively little data on how these
fluctuations affect pressure. The majority of research on water
supplies has been conducted for determining the quantity of water end
users consume during certain periods of time. In most cases, this
research is focused on domestic users because they are easily
categorized by type of residence and number of individuals. Commercial
and industrial users are harder to categorize due to the many uses of
water other than domestic.
determined that there was insufficient data to provide recommendations
at this time regarding water supply adjustments. The reasons: 1) a lack
of data associating flow rates and available pressure, 2) insufficient
data to provide meaningful comparisons between regions and within
specific regions, 3) a lack of data for all identified variables, and 4)
data was not limited to a single variable or discrete number of
variables that would allow for development of adjustment factors.
WHAT SHOULD DESIGNERS DO NOW...
an adopted adjustment factor, the system designer and plans reviewer
should contact water authorities when performing flow tests to determine
the best time to conduct the test during normal or high demands. For
tests that have already been conducted, designers can ask how the flow
conditions observed during the test compare to normal or high system
flow conditions. This would meet the current recommendations of NFPA 291
for conducting tests during a period of ordinary demand.
water authorities provide modeled hydrant flow for use in sprinkler
system designs, they should be asked how well the hydraulic models are
calibrated (how the modeled hydrant flow tests compare to actual hydrant
flow tests) and reduce the modeled hydrant flow test by the percentage
difference they observed in their calibration (e.g., 10%, 20%, 25%,
When designers are unable to
contact the water authority, they should perform several flow tests at
different times of the day to determine when the period of normal or
high demand occurs, or the designers can design the sprinkler system
with a margin of safety with the intent to account for variation in the
water supply. Design practices such as reducing pipe sizes that increase
demand pressures should be minimized in areas where the variation in
the water supply is not known.
it is the designer’s responsibility to understand the characteristics
of the water supply they are using to supply their fire protection
system. Recent research has identified major factors that contribute to
these variations; however, insufficient data is currently available to
quantify these factors for the development of an overall adjustment
Joseph E. Kurry and Mark Hopkins are with JENSEN HUGHES.