Backflow devices have a relatively long history with water-based fire protection systems. For many years, the separation between the stagnant water in a sprinkler system and a potable water supply was accomplished by a single check valve on the fire protection system's water supply line. However, in the 1970s, many water supply purveyors and the plumbing codes began requiring backflow prevention devices on those water supply lines. 

Two standards published by the National Fire Protection Association prescribe the requirements for the installation of backflow prevention assemblies (NFPA 13, Standard for the Installation of Sprinkler Systems) and the ongoing inspection, testing and maintenance of them after installation (NFPA 25, Standard for the Inspection, Testing and Maintenance of Water-Based Fire Protection Systems).

The fire sprinkler industry has had concerns about using backflow preventers in fire protection systems since they were first required. These devices are normally used in applications where water flows through them regularly, such as irrigation systems. However, water doesn’t flow through a fire protection system unless there’s a fire event or unless flow testing is being performed. 

Forward-Flow Test

Because the backflow preventers are in a static position almost 100 percent of the time, bonding may occur between the check valve disc and the seat, resulting in a delayed or partial opening of the check valve during a fire event. This creates an impairment of the fire protection system. Another concern is a buildup of corrosion and mineral deposits within the device. 

As a result of these concerns, special tests are required by NFPA 25 when backflow preventers are installed on water-based fire protection systems, including a forward-flow test. The purpose of this test is to ensure that water can flow freely from the public water supply through the backflow preventer to the fire protection system.

When a standard check valve was used to prevent backflow, the inspection and maintenance requirements in NFPA 25 were virtually the same as they are today. The 1992 edition of NFPA 25, which is the first edition of that standard, required that check valves be internally inspected once every five years. The internal components such as the clapper and clapper seat were to be cleaned, repaired or replaced when necessary to meet the manufacturer’s instructions. There wasn’t a forward flow or any other test required for check valves.

In that same edition of NFPA 25, backflow prevention assemblies had very different inspection, testing and maintenance requirements. A reduced pressure type had to be inspected weekly to ensure it wasn’t constantly discharging water through the differential sensing-line relief port. 

There were two requirements for testing all types of backflow prevention assemblies. First, each one had to be tested annually for backflow capabilities. That test usually required the inspector to be certified by the manufacturer to meet the local plumbing code requirements. 

Second, a forward-flow test through the assembly had to be conducted annually at the water-based fire protection system's design flow rate, including hose streams. The flow rate (gpm), the inlet pressure (psi) and the outlet pressure (psi) had to be measured so the friction loss across the backflow device could be calculated. The friction loss was then compared to the manufacturer’s specifications. 

The appendix of NFPA 25 explained that the forward-flow test could be accomplished by flowing water through a yard hydrant, a fire pump test header or standpipe hose connections, through the fire department connection after reversing the check valve, or through a small hand hose or other outlets that will discharge sufficient flow. 

The intent was to measure the flow and pressures and then calculate the friction loss across the device to ensure it wasn’t impeding the water supply to the fire protection system.

The Exception

There was an exception provided that when sufficient size outlets weren’t available to flow the system demand, the test was to be conducted at the maximum flow rate possible. This exception has carried through in all subsequent editions, which today is the cause of some confusion. 

The 1994 edition of NFPA 13 recognized that when a new sprinkler system was installed, it must include a “means” downstream of a backflow preventer to perform the forward-flow test required annually by NFPA 25. Suggestions were made of what the means could be, including a bypass around the fire department’s connection check valve. 

However, no specific means has ever been described in NFPA 13, even though several recommendations were made to require 2 1/2-inch hose valves to be installed for testing purposes. The unfortunate reality is that installing contractors rarely include a realistic means to perform the forward-flow test. 

It’s up to the ingenuity of the person performing the annual flow test to determine the means available to actually measure the flow. 

Testing Confusion

One method commonly used before exceptions were added to the standard was to drain the system down, reverse the check valve on the fire department connection and flow out through the FDC.

Through the subsequent editions of NFPA 25, the inspection, testing and maintenance requirements for backflow preventers gradually changed to the point where today, there is a great deal of confusion about what is required and how to perform the forward-flow test. 

In the 1998 edition, the requirement to calculate the friction loss across the device was deleted. In that edition, it was suggested, as it was in NFPA 13, that a bypass around the fire department connection with a normally closed control valve or a closed loop around the backflow assembly with a flow meter could be used. Both of these suggestions are relatively expensive to install if not included in the original system installation. 

In the 2011 edition, the requirement for the backflow performance test was deleted. Since this test was to comply with the plumbing code, it was dropped from the fire protection standard. In the 2014 edition, a requirement was added for an internal inspection of backflow preventers once every five years, except if it was performed during annual maintenance for the device.

The current 2020 edition of NFPA 25 includes the following requirements related to the periodic inspection, testing and maintenance of backflow preventers:

• A weekly inspection is required for reduced-pressure backflow assemblies to ensure the differential-sensing valve relief port isn’t continuously discharging.

• A quarterly main drain test downstream of the backflow preventer must be performed when it’s installed in the sole water supply for the water-based fire protection system.

• An internal inspection must be performed at least once every five years to ensure the internal components are in good condition, move freely and operate correctly.

• Maintenance for a backflow preventer is to be performed by a qualified person, follow the manufacturer’s instructions and comply with the authority having jurisdiction’s and local jurisdiction’s policies.

These requirements are fairly straightforward; compliance should be achievable without too much confusion or second-guessing.

This edition also includes the forward-flow testing requirement and allows it to be conducted in one of two ways:

1. By flowing the minimum flow rate of the system demand or more through the means provided; or

2. By flowing the maximum flow rate possible when a means hasn’t been provided.

Questions are constantly being raised on how to perform this test on a sprinkler system that doesn’t include a test header, as a fire pump does, or hose connections, as a standpipe system. Today, the forward-flow test should be easy to comply with, especially for newer sprinkler systems installed following the 1999 and later editions of NFPA 13. The means to perform the test at the system demand flow rate is required to be installed.

It’s the exception when a means hasn’t been provided that confuses. Section 13.7.2.3 states, “Where connections do not permit verification of the [forward-flow] test at the minimum flow rate of system demand, tests shall be conducted at the maximum flow rate possible.” 

In other words, if the means to flow the system demand through the backflow preventer isn’t provided, the forward-flow test should be conducted by flowing as much water as possible through the device. For a wet pipe system, this would include opening the main drain valve and the inspector’s test connection to generate the maximum flow rate possible.

By opening the inspector’s test connection, there may be another 50 gpm flowing in addition to what’s flowing through the main drain. However, these combined flows would typically not be equal to or greater than the system demand flow rate. 

For dry pipe systems, this would include opening only the main drain valve. Where multiple sprinkler systems are downstream of a backflow preventer, all drain valves and test connections should be opened simultaneously to generate the maximum flow rate possible. Does this exception meet the intent of the forward-flow test, or should additional effort be made to flow and measure the system demand rate?

This column isn’t going to resolve this issue. In my opinion, all reasonable efforts should be made to flow the system demand rate through the backflow preventer annually. The quarterly main drain test will lift the check valves but don’t prove they will open fully should the system demand rate flow through the backflow preventer in a fire event. 

Perhaps the next edition of NFPA 25, which is currently in a revision cycle, will finally clarify this requirement.