How Do Backflow Prevention Assemblies Work?

In order to repair any assembly, RP, DC, PVB or SVB, it is important that the repair technician first understand how the backflow prevention assemblies are supposed to work, so that when they are not working, the problem can be properly identified. The purpose of the repair process is to return the assembly back into its original factory specifications.

When we are field testing any of these assemblies, we are making a diagnostic analysis at one point in space and time. From this data, we cannot say how the backflow prevention assembly worked yesterday or if it will work tomorrow, only what it is doing right now. When we perform the field test, we are generating data on the workings of the assembly, which we must compare to the minimum acceptable performance data as established in our accepted test procedure. For example, two (2.0) PSID is the minimum acceptable relief valve opening point, but few assemblies are designed to open at 2.0 PSID. For this reason, when we get a performance value from a field test, whether it be a relief valve opening point, or a check valve value, we must understand what is happening in the backflow prevention assembly.

Reduced Pressure Principle Backflow Prevention Assembly (RP)

To understand what happens inside an RP, let us flow water through a generic backflow prevention assembly. An RP consists of inlet and outlet shut-off valves, four properly located test cocks, a first and second check valve component, and a relief valve component. Let us hook up our RP to a water source, which delivers 100 PSI, and begin to pressurize the backflow prevention assembly.

Relief Valve

As the inlet shut-off is opened, water enters the upstream side of the backflow prevention assembly body ahead of the 1st check valve. Once in this area, the water enters a relief valve sensing line. Some sensing lines are external hose or pipe, and some utilize an internal passageway. The water travels through this sensing line to the elastic element in the relief valve. The elastic element is either a diaphragm, or a rolling diaphragm. This pressure will build up on the high pressure side of the elastic element, which will deflect and cause the relief valve stem to compress the relief valve spring, and move the relief valve disc to seal against the relief valve seat. The function of the relief valve spring is to constantly try to open the relief valve.

If the pressure downstream of the 1st check rises to where it is a minimum of 2.0 PSI less than the inlet pressure upstream of the 1st check, then the 2.0 PSI loading from the relief valve spring would cause the relief valve to open. In a properly working relief valve, the opening point can be anywhere from 2.0 - 5.0 PSID depending on the manufacturer, model and size. Once the water pressure has closed the relief valve, then the pressure will increase to the next point, which will cause the first check to open, and allow the water to travel past the first check valve.

1st Check Valve

The water pressure will be reduced by the amount of pressure it takes to open the first check spring loading. The 1st check of an RP can have a loading of anywhere between 5-15 PSI depending on the make, model and size. In our generic RP, we will assume our relief valve spring generates a 2.0 PSID and the first check spring has a 10.0 PSI Loading.

Once the 1st check opens, water will travel past and pressurize the area between the 1st and 2nd checks. When this area is pressurized, it will also pressurize the low pressure side of our relief valve. The higher inlet pressure (100 PSI) is placed on the high pressure side of the elastic element in the relief valve, and the lower pressure past the 1st check (90 PSI) is placed against the low pressure side of the elastic element.

2nd Check Valve

Once this area between the two checks and the low pressure side of the relief valve is pressurized, the pressure will now cause the 2nd check to open. The loading of the 2nd check spring will be anywhere between 1-5 PSI depending on the make, model, and size. The pressure after the 2nd check will be reduced by the amount of pressure it takes to open the 2nd check. For our illustration, we will give the 2nd check a 5.0 PSI loading.

The upstream pressure (100PSI) of the RP in our illustration is reduced by the combined pressure load of the 1st and 2nd check spring (100 - 10 - 5 = 85PSI), producing our downstream pressure of 85 PSI into our piping system. Don't forget that the relief valve spring is continually trying to open the relief valve, which the inlet pressure is keeping closed, and that a properly working relief valve can only open when the pressure downstream of the 1st check plus the relief valve spring load is greater than the upstream pressure to our RP. In our illustration, the pressure past the 1st check (90 PSI) must increase to 98.1 PSI, so that the value of the relief valve opening point (2.0 PSID) and the pressure after the 1st check (90 PSI) is now greater (98.1 + 2.0 > 100PSI) than the inlet pressure.

During the normal flow of water through a properly working RP backflow prevention assembly, the relief valve will be pressured closed, and the check valves will modulate between an opened and closed position to fill the water demand of the plumbing system. The two check valves (1st & 2nd Check) will have different spring loadings. The check valves in an RP will open when something downstream of the RP in the piping system opens a water-using fixture, and as the water begins to flow to that fixture, the pressure drops. The pressure upstream of the fixture begins to drop as the water flows through the fixture. Because of this flow of water, the pressure upstream of the RP is now great enough to cause the check valve to open and flow water to the fixture demanding water. Check valves only open enough to fill the demand for water; they do not open fully, but usually modulate to fill the demand for water.

Backpressure Condition

Backflow is the hydraulic condition that can cause an RP to stop working in the described normal flow pattern. Backflow can happen by either backpressure and or backsiphonage. Backpressure is a condition where a greater pressure is generated on the downstream than the upstream side of the assembly. This condition can happen for many reasons, pumps, thermal expansion, etc.

Let us assume we have a proper working RP, and we apply backpressure to the outlet side of our RP. In our illustration, the normal upstream pressure is 100 PSI and the downstream pressure is 85 PSI. If the starting downstream pressure (85 PSI) increases to105PSI for example, and the second check is working properly, the 2nd check closes and keeps the 105PSI pressure from migrating into the area (90 PSI) between the 2 check valves.

Even if we have a working 2nd check, and backpressure is applied, we can get a discharge from our relief valve. A condition called disc compression can cause discharge from a properly working RP. When backpressure occurs, this increase in pressure placed on the downstream side of the 2nd check causes the 2nd check disc to embed farther into the 2nd check seat. The volume of water in the body between the 2 checks is being squeezed as the 2nd check disc embeds farther into the seat. Water is a not a compressible fluid in these pressure ranges, so this squeezing of this water causes an increase in pressure in the area between the 1st and 2nd check valves. If this increase in the pressure between the 2 checks, which started at 90 PSI in our illustration, increases to the point where it is greater than the inlet pressure minus the relief valve spring loading (100 - 2.0 = 98), the relief valve will open.

In our field test procedures, when we perform the 2nd check test of an RP, we are simulating a backpressure condition by bringing the higher inlet pressure (100 PSI) around to test cock 4 (85 PSI). If you remember from your field test procedures, when an apparent 2nd check failure is observed, you are required to open your low side bleed valve on your test kit. This will draw the elevated pressure from the area between the 2 check valves, while the second check disc stays embedded into the 2nd check seat from the applied backpressure. When the low bleed is opened, you are reestablishing the pressure in the area between the 2 checks back to its normal pressure of 90 PSI while the elevated 100 PSI is maintained after the second check.

Disc compression is one of the most common errors made by a backflow prevention assembly tester when performing a field test. Once the relief valve discharges when testing the second check, you must open the low bleed one more time to determine if the 2nd check is actually working or not. A disc compression scenario may occur, and the tester may incorrectly assume the 2nd check is not working.

Let us see what happens when we apply backpressure to a non-working 2nd check. Once the pressure begins to increase on the outlet of our assembly, the 2nd check cannot maintain the separation of pressures between the inlet and outlet of the 2nd check, and the pressure will equalize on both sides. As the pressure increase begins (from 85 PSI in our illustration) the area between the 2 checks will also increase. Remember that the area before the 2nd check is where our low pressure is applied to the low pressure side of our relief valve elastic element. As the pressure increases above our starting pressure of 85 PSI, and goes to the point equal to the inlet pressure minus the relief valve opening (100 PSI - 2.0 PSI = 98) of 98 PSI, the relief valve will open.

Backsiphonage Condition

Backsiphonage is a condition that causes a sub-atmospheric pressure to be applied to the upstream side of the assembly. Backsiphonage can happen for several reasons; one of the more common is excessive water demand in the distribution system.

Let's examine the effect of backsiphonage on our RP and see how it functions. When the inlet pressure to our assembly (100 PSI in our illustration) goes down to sub-atmospheric, or negative, the 100 PSI is reduced to a negative pressure. The pressure at the inlet of the RP is what keeps the relief valve closed. When the pressure at the high pressure side of the elastic element in the relief valve is reduced to a negative, the relief valve will open because of the relief valve spring load, and any pressure remaining in the area between the 2 check valves is now applied to the low pressure side of the relief valve elastic element.

The relief valve of an RP can only open for the two backflow conditions of backpressure or backsiphonage or the simulation of these two conditions. One of the more common simulations of backpressure happens when there is a pressure fluctuation at the inlet of the RP. If there is no flow going through the assembly, and then the upstream pressure drops quickly from 100 PSI down to 80 PSI, this can cause the relief valve to open. This would happen because there would be higher pressure on the low pressure side of the relief valve elastic element versus the high pressure side (high pressure side would be 80 PSI and the low pressure side would be 90 PSI plus the relief valve spring loading).

If a water hammer condition happens on the downside of a properly working RP because of a quick closing solenoid, this increase in pressure would create a backpressure condition, which could cause the relief valve to open by disc compression.

Just because a relief valve discharges water, does not always mean the RP is not working. Pressure fluctuations can simulate conditions that can lead a person to assume the assembly is not working properly. To be sure whether the assembly is working or not, a test kit must be attached and proper test procedures applied to determine the working condition of the assembly.

Faulty Component Diagnosis

Now let us look at how an RP reacts when components of the assembly are not working properly and how we can diagnose the condition. Let's assume we perform a field test on an RP and generate a 1.5 relief valve opening point. Does this mean the assembly is "leaking" or will not prevent backflow? The answer is probably no. We know from our tester training that 2.0PSID is the minimum acceptable relief valve opening point. If the relief valve opens at 1.5PSID, the relief valve will open and keep the pressure in the area between the two checks 1.5 PSI lower than the upstream pressure, if subjected to a backflow condition. An RP with a 1.5 PSID relief valve opening point will still prevent backflow, but does it at a value lower than the minimum accepted value of 2.0 PSID. For the assembly to perform optimally, it must operate at or above this minimum standard, in this case 2.0 PSID.

Low Relief Valve Opening Point

The cause of a relief valve opening below the 2.0 PSID minimum can vary greatly between different models. Usually, the incorrect assumption is made that a spring has worn out and that is why the relief valve will not open. The most common cause of low relief valve opening points is a restriction on the travel of the relief valve stem mechanism. Either the guide of the relief valve becomes damaged, or a scale or corrosion will cause the guide to not travel optimally leading to a low relief valve opening point.

High Relief Valve Opening Point

What happens when our field test data presents an excessively high relief valve opening point, something above 5.0 PSID? A high relief valve opening point can happen for different reasons depending on the make, model and size. The most common cause of a high relief valve opening point is when the relief valve disc does not completely embed itself into the relief valve seat, usually because the relief valve stem assembly is not traveling its full length; for example, if a rolling diaphragm in the relief valve is pinched or twisted, it will restrict the relief valve from traveling its full designed length. If the stem does not travel its full length, the relief valve disc cannot fully embed into the relief valve seat. If this disc is just barely touching the relief valve seat, and not fully embedded, then the relief valve opening point will be excessively high.

Faulty First Check Valve

Let us see what happens when the check valves are not working properly. Let us talk a little about the first check. In this example we show an upstream pressure of 100 PSI. The pressure downstream of the first check shows us 90 PSI, which means we have an 10 PSID across the 1st check valve. This is the load the first check is generating on a properly working first check. If the first check was completely fouled and there was no differential produced, that means we would have 100 PSI before and after the first check (0 PSID); then the relief valve spring would cause the relief valve to stay open.

The first check rarely fails where there is no differential. The usual case is that instead of a 10 PSID, as shown in our example, the differential begins to fall as the first check begins to wear out. Let us assume we know our relief valve has a 2.1 PSID opening point. Let's add further that our first check is starting to degrade and it can only generate a 2 PSID. In other words our upstream pressure is still 100 PSI and the downstream pressure of the first check is 98 PSI and we know we have a 2.1 relief valve opening point, what would happen to our relief valve? The relief valve would open up and begin to discharge. If we have a 100 PSI inlet pressure and a pressure of 98 PSI after the first check, you can see where the 98 PSI along with the 2.1 PSI from the relief valve spring loading would cause the diaphragm to move, causing the relief valve to open, because there is a greater pressure on the downstream side of the relief valve diaphragm (98 +2.1 =100.1 PSI) than on the upstream side (100 PSI).

Some administrative authorities require the loading on the first check to have a minimum of 3.0 PSID higher value than the relief valve opening. By having a buffer greater than 3.0 PSID, this would help minimize relief valve discharge from a small pressure fluctuation in a static condition. This would mean that if our relief valve opening point is 2.1 PSID than we would have to have a first check loading of at least 5.1 PSID to pass the field test. If a 3.0 PSID buffer was not required in your area, then any first check value greater (above 2.1 PSID) than the relief valve opening point would keep the relief valve closed and would be a passing check value.

The cause of check failure tends to be due to the failure of the disc to seal against the check seat easily. Many times the check spring is blamed for a check failure but this is usually not true. The more common causes are dirt and debris on the disc, disc degradation where the disc will not seal, or a check guide restricting the travel of the check component.

Faulty Second Check Valve

The criteria for the workings of the second check, like the first check, must maintain a higher pressure upstream of the check than the downstream pressure. This differential is established by the spring loading of the second check spring, which is designed to be a minimum of 1.0 PSID.

The test procedure for the second check is different than the first check. The 2nd check test is a backpressure test, while the 1st check is a direction of flow test. In our field test of the 2nd check, we take the higher inlet pressure form test cock #2 upstream of the first check (100 PSI), and with needle valves and hoses, place it into our number four test cock (85 PSI) causing the pressure on the downstream side of the second check to rise until it is higher than the upstream side of the second check. When the field test is performed on a working 2nd check, the check disc will be embedded farther into the check seat. This can cause a condition known as disc compression as previously mentioned. When the second check fails, the higher pressure would go past the leaking second check into the area between the two checks. As the pressure in this area increases, the relief valve senses the differential. When the pressure in the area between the two checks increased to 98.0 PSI (relief valve opening point 2.1 PSID), then the diaphragm would move, causing the relief valve to open. The causes of failure on a second check are similar to the first check.

In conclusion, the field test is the way we generate the data needed to determine which part of the assembly is performing below the accepted minimal standard. When the numbers fall below the minimum standards established by the accepted test procedure, a repair must be facilitated to bring the working condition of the assembly above the minimum standards and back to its original factory working specifications. The generation of accurate data is very important and this means using an accurate test kit and proper test procedures and techniques to ensure that the data we generate properly reflects the working condition of the assembly.

Double Check Valve Assembly - DC

A DC is simply two approved independently operating check valves that can hold a minimum of 1 PSI in the direction of flow with the outlet of the check valve open to atmosphere. These checks must be located between an inlet and outlet shut-off and have 4 properly located test cocks. The check valves in a DC must hold a minimum pressure (1.0 PSI minimum) in the direction of flow. The two check valves (1st & 2nd Check) will have similar spring loadings.

The check valves in a DC will open when something downstream of the DC in the piping system opens a water-using fixture, and as the water begins to flow to that fixture, the pressure drops. The pressure upstream of the fixture begins to drop as the water flows through the fixture. Because of this flow of water, the pressure upstream of the DC is now great enough to cause the check valve to open and flow water to the fixture demanding water. Check valves only open enough to fill the demand for water; they do not open fully but usually modulate to fill the demand for water.

If a check valve is holding less than 1.0 PSI, for example, 0.5 PSI, testers have been known to incorrectly state, "The check valve is leaking." This leads some people to believe that this check valve would not stop a backflow condition. In fact, the check valve is not leaking at 0.5 PSI because it is still sealing off the area upstream and downstream of the check valve with a 0.5 PSI loading; however, the check valve is performing below the minimum criteria as established in the test procedure (1.0 PSI).

The minimum criteria in a test procedure is set at a point that will trigger a repair before the assembly can degrade to the point where it cannot prevent backflow (0.0 PSI). As long as our check valve has a positive loading, it can prevent backflow, but only when it is above 1.0 PSI does it meet the minimum criteria as established in the test procedure. So once the test procedure generates data that the check is maintaining less than 1.0 PSI, we must repair the check valve and return it to its original working specifications.

Conditions that can cause a check to perform below its optimum level are many. The main cause of check failure is due to the failure of the disc to seal with adequate force against the check seat. Many times the check spring is blamed for this lack of force but this is usually not true. The more common causes of failure are dirt and debris between the disc and seat. Another common problem is disc degradation where the disc will not seal tightly against the check seat. The third common cause is a restriction in the travel of the guide, limiting the movement of the check valve and prohibiting it from sealing properly.

There is a variation of a DC called a (DCDA). This is a double check created for fire sprinkler applications. A DCDA consists of an approved DC with a bypass arrangement that consists of a by-pass water meter and an approved by-pass DC. This by-pass is piped around the mainline DC. The purpose of this by-pass arrangement is to detect and register the first 3 gallons per minute (GPM) of flow across the backflow preventer into the fire system.

Many testers think a DCDA is simply any small by-pass DC piped around any main line DC with a meter attached, and because the by-pass DC is smaller, the first flow will go through it. This is not true. In order for the by-pass to detect and register 3 GPM, the two DC's and the water meter must be engineered so that the larger assembly will have a slightly higher differential at the low flow condition of 0-3 GPM. This will ensure that the first 3 GPM travel through and are registered by the water meter in the by-pass. Then, if the fire system demands more than 3 GPM, the main line assembly will open up and flow will commence up to the designed flow requirements of the system.

An inexperienced installer may install a mainline DC and pipe in a small by-pass that looks similar to the DCDA. These unapproved DCDA's cannot guarantee that they will detect this first 3 GPM because they are not factory engineered assemblies with the proper pressure differentials, but rather a collection of 2 DC's and a water meter assembled to look like a DCDA.

Most fire protection systems do not have a mainline water meter at their point of service from the water purveyor, and that is why it is important to detect this low flow of water. Because a fire system is an emergency connection, water purveyors do not want the expense or extra flow loss of going through a full size water meter. Because a fire system is considered an emergency connection, there should be no flow to detect across a mainline water meter anyway. Water purveyors use this bypass to ensure that water users with fire systems do not flow water through this emergency connection, and with the ability of the bypass to detect small flows, they can also detect if there are any small leaks that may be underground and out of sight.

There is also a Reduced Pressure Principle Detector Assembly (RPDA). This has a main line RP with a bypass arrangement containing a smaller RP and detector meter to register the 1st 3 GPM of flow. These RPDA's are used on high hazard fire systems, and the DCDA is used for low hazard fire systems, when the detector function is needed.

The check valves of a DCDA are similar to the DC and the repair process will be similar. In many cases, the spring loading of the mainline assembly or the by-pass may be different from the standard DC, but the repair procedures and the test procedures to diagnose its workings are the same. Before we can repair any assembly, it is important to have correct data on the workings of the assembly to be sure we know what we are fixing, and just as importantly, that it really does need to be repaired.

Pressure Vacuum Breaker Assembly (PVB) and the Spill Resistant Pressure Vacuum Breaker (SVB)

The PVB consists of an inlet and outlet shut off, two test cocks, a check valve, and an air inlet component. The normal flow of water goes from the upstream side of the body into the check valve. The check valve (which is spring biased closed) is designed to hold 1 PSI in the direction of flow, similar to the check in a DC.

The check valve opens, and water travels past the check valve, and causes an air inlet poppet (spring biased open) to travel up an air inlet guide, compressing an air inlet loading (it's not always a spring), which is designed to generate a load of at least 1 PSI. The air inlet is kept closed by the normal water pressure in the piping system and is designed to open when the force from the air inlet (1.0 PSI Minimum) is greater than the water pressure in the area downstream of the check valve. The PVB is designed to prevent backflow from backsiphonage only, and must be installed 12" above the highest point of use or piping on the downstream of the assembly.

PVB Check Valves

Conditions that can cause the check in a PVB to perform below its optimum level are similar to the check failures in an RP or DC. In addition, there is a common cause of failure unique to the PVB and SVB, which has to do with the alignment of the check spring. Many models require the spring to be installed with a spring retainer that, if not properly installed, will exert a side pressure on the spring, preventing it from delivering the proper load to the check valve.

Air Inlet

Sometimes the air inlet will fail, in which it will not fully unseat itself when the water pressure in the body past the check valve is below 1.0 PSI. This can occur when the air inlet disc adheres to the air inlet seat, caused by temperature or water quality conditions.

In other cases, the canopy that covers the bonnet is missing, which can also cause direct sunlight onto the air inlet, also causing a problem with deterioration from the ultraviolet rays of the sun. Lastly, On some models of PVB's, the air inlet spring can easily be removed or inserted in such a way as to lower its loading below the 1.0 PSI minimum requirement. There is one brand of PVB that does not use a mechanical spring in the usual sense but rather a fold of rubber on the poppet that generates the load, and if you are not familiar with this brand, you could erroneously assume the spring is missing.

Sometimes the air inlet poppet will not seal on the air inlet seat completely, and will leak. This unwanted discharge from the air inlet can be caused by several reasons. The usual is when some dirt or debris is located between the air inlet poppet disc and the air inlet seat. If the disc becomes damaged from this debris or becomes worn for other reasons, it could inhibit its ability to seal. Another cause of leakage can happen if the air inlet guide is damaged in such a way as to not allow the air inlet poppet to seat squarely on the air inlet seat.

SVB

There is a variation of the PVB called an SVB. The SVB has an inlet and outlet shut off, a check valve and an air inlet valve, a single test cock, and a bleed screw. The SVB performs similarly to the PVB except when initially pressurized.

The normal path of water for a PVB is for water to enter the body, then open the check valve with a minimum 1.0 PSI loading, proceed past the check valve, and seal the air inlet with a minimum 1.0 PSI loading. The SVB is a little different. The check and air inlet have the same 1.0 PSI check and air inlet minimum loading requirements, but the order of the operation of the components is different.

When water first enters the SVB, instead of causing the check valve to open first, as in a PVB, the air inlet closes before the check valve opens. This is accomplished by the air inlet having a lighter loading (1.0 PSI minimum) than the check valve (1.0 PSI minimum). Even though they have the same minimum loading requirement, there will always be a pressure differential between the load of the check and the air inlet, with the check valve having a higher loading than the air inlet.

In this way, water seeking the path of least resistance will close the air inlet before it can open the check valve, thus minimizing any discharge (spill resistance) from the air inlet. In the SVB, water does not have to travel past the check valve to pressurize the air inlet as it does in the PVB. For this reason, the SVB will not discharge from the air inlet on initial start up. Once the SVB is pressurized, the SVB will perform similar to a PVB. The causes of failure of a SVB are similar to those of a PVB as discussed above.

In order to determine how well a backflow prevention assembly is working, it is important to understand how it operates when it is in a working condition and a non-working condition. It is important that we can diagnose various conditions by performing an accurate field test with an accurate test kit to generate accurate data to make a true assessment of the workings of our assemblies.