I’ve seen wringing of hands and gnashing of teeth whenever the subject of pumps and reduced pressure principle backflow preventers (RPs) is broached. Why would anyone want to install both in the same system? How did we get here? Flicker your fingers before your face a la Dana Carvey from Saturday Night Live’s Wayne’s World (indicating the start of a dream sequence). Sound effects are optional. Hopefully someone is reading this article aloud to you at this point.
You are in ancient Sicily witnessing a conversation between a frantic farmer and the great engineer and philosopher Archimedes, who wipes a bead of sweat from his nose as the sun beats down on the parched earth.
“Master, the barley will die soon if we don’t get it watered. There will be no bread for our wine and cheese parties, and no porridge for the children. If we can’t get the water from the river to the crops soon, we’re totally screwed.”
Archimedes strokes his long beard. “Screwed, eh?”
Shortly thereafter, Archimedes delivers what is perhaps his most famous invention, the screw pump, and all is well.
“You’ll want to leave an air gap there at the top, so the sheep dung you put along the rows doesn’t run back into our drinking water in the river.”
Well, that’s how I imagine it all went down anyway.
Flash forward a few millennia. Several World War II era backflow incidents spawned the formation of the E.C. Service Company and the development of the first RPs. Ever since, pumps have been used in combination with them. Why? Simply put, to boost pressure when the available pressure is not adequate to get the job done. Where will you find pumps? In building intake lines, wells, fire protection systems, hot water circulation systems, sump pits and irrigation systems.
There are a few technical terms you should be familiar with when talking about pumps. The first is head pressure. This is usually stated in a unit of length, like feet, and is simply pressure (force/area) divided by the pumped fluid’s specific weight (weight/volume). The next term is net positive suction head required (NPSHR). This is the minimum fluid energy required at the pump inlet to make it work. NPSHR is determined by the pump manufacturer, is flow-dependent, and cannot be adjusted in the field; it is what it is. Complementing NPSHR is NPSHA, or the net positive suction head available. As the name implies, this is how much energy is actually available at the suction side of the pump. Contributors to this energy include atmospheric pressure and the potential energy of the fluid due to its height above the ultimate point of delivery. Detractors from NPSHA include frictional and vapor pressure losses. Cavitation is another term you may hear with regard to pumps. Pump cavitation occurs when vapor bubbles form and subsequently implode within the pump. This occurs when the absolute pressure on the liquid falls below its vapor pressure. Cavitation will occur when the NPSHA is less than the NPSHR.
On occasion, you may come across a pump curve. Although it may appear daunting at first glance, a pump curve is actually a concise way to share lots of information about a pump.
Typically, the left vertical axis indicates head pressure, the right vertical axis monitors power and NPSHR, and the horizontal axis indicates flow. As expected, the head flow curve shows that the pump can deliver higher pressures at lower flow rates, and lower pressures at higher flow rates. Iso-efficiency curves are pressure and flow combinations that occur at the same efficiencies.
What is pump efficiency? It is the power output of the pump, divided by the power required to run it. Pump manufacturers will identify a best efficiency point (BEP), indicating the “sweet spot” where the pump runs most efficiently. Most in the industry pronounce BEP as “beep,” like half the utterance of one of my favorite cartoon characters.
Pumps used with RPs typically fall into two broad categories: positive displacement and centrifugal. Positive displacement pumps entrap a fixed amount of liquid, displace it, and thereby force flow. These pumps produce the same flow at a given speed, regardless of discharge pressure, because the volume is constant through each cycle of operation. Archimedes’ screw pump falls into this category, as does a well hand pump. What you are more likely to see in combination with an RP is a more modern version.
A positive displacement pump has an expanding cavity on the suction side of the pump and a decreasing cavity on the discharge side. Water is allowed to flow into the pump as the cavity on the suction side expands and the liquid is forced out of the discharge as the cavity collapses.
What problems might you encounter with a positive displacement pump? The off-on nature of its pumping cycle, even with the double-acting type that supplies fluid on its suction and delivery strokes, means there will be inherent pressure fluctuations. Tight internal clearances make PD pumps prone to velocity erosion wear, especially when they’re operated at high speeds. Then there is the ever-present threat of cavitation.
You are more likely to encounter a centrifugal pump used in combination with an RP. In this type of pump, water is sucked into the eye end of a rotating impeller. Changing areas inside the casing (volute) pressurizes the discharge flow.
What are the concerns with a centrifugal pump? They may overheat, especially at lower flow rates, they must be filled (primed) to work properly, they use long rotating shafts that are sometimes prone to leaking and, like PD pumps, they can cavitate.
To be clear: when operated per manufacturers’ recommendations, both types of pumps can deliver years of trouble free service. So why do people get nervous when pumps and RPs are used in the same system? Perhaps because both products prefer a smooth, non-fluctuating pressure source, and each
can disturb the other.
When in the field, you will see RPs installed both before and after a pump. So, which orientation is correct? When it comes to fire service applications, National Fire Protection Association (NFPA) 20’s Technical Committee has a long standing recommendation that NO backflow preventer be installed on a system that uses a fire pump, but if required, they prefer them on the downstream (system) side.
Some AHJs prefer the RP to be on the upstream side of the pump suction flange because they consider the pump to be a cross-connection, especially if the pump is used as part of a chemical injection system. When mandated by an AHJ, NFPA requires the RP to be a minimum of 10 pipe diameters upstream of the pump inlet.
Regardless of the pump type used, or where it is situated in the system, the pump/RP combination can cause problems. When the RP is installed downstream of the pump, any debris collected in the pump can foul the RP’s checks and relief valve during pump start-up. This is especially true in well applications. A quick pump shut down can also suddenly reduce the RP’s inlet pressure, causing unintended relief valve discharge. Frequent pump cycling can cause excessive wear on the RP’s components. When the RP is installed upstream of the pump and the pump shuts down quickly, a water hammer shock wave will be generated, which will move back toward the RP. The result? Nuisance relief valve discharge and excessive wear on the second check. Most of what can be done to mitigate these unintended consequences is out of the hands of an inspector, tester, and manufacturer, and must be initiated by the property owner.
If you get an inquiry about a pump / RP combination causing a problem, check the following:
- Is a variable frequency drive (VFD) being employed for pump start up and shut down? VFDs provide “soft” starts and stops, and lessen the intensity of pressure fluctuations and water hammer.
- Are flow straighteners being used? These devices (imagine a bundle of straws contained within a straight run of pipe) smooth the flow and cut down on turbulence.
- Is cavitation occurring? This is usually happening when you receive a call stating, “What do you mean you can’t hear me? I said it sounds like rocks are running through my pump! Can’t you hear that!?” There are a variety of causes for cavitation, but the most likely are the water is too hot, the pump is running too fast, or the suction lift (depth from which the water is pumped) is too high. All of these conditions lead to the NPSHA being less than NPSHR. A different pump may be necessary to lower the NPSHR value.
- Are straight runs of pipe (minimum of 10 pipe diameters if the RP is in front of the pump and five pipe diameters if the RP is installed after the pump) being used? You may be surprised at how many times you will find an RP attached directly to a pump! There is way too much turbulence and pressure fluctuation for this configuration to be effective.
- Is a spring loaded check valve installed between the RP and the pump? This will dampen the water hammer shock wave for an upstream RP installation and prevent unnecessary relief valve discharge for a downstream RP installation. A soft or hard seated design can be used. Soft seated checks work (dampen) slightly better, but will wear much more quickly than hard seated checks.
One final word of caution: whether you are physically adjusting something in a flow line or advising someone who is, please take care not to “dead head” a pump. Dead heading has nothing to do with humming Truckin’ while conducting a backflow test; it is a situation that occurs when a pump’s discharge is closed, either due to a blockage in the line or a closed valve. The pump will go to its maximum shut-off head and the water will recirculate within the pump and eventually vaporize, causing pump damage at minimum. Dead headed pumps have been known to explode and cause serious injuries. Field test procedures can require the first or second RP shut-off valves to be closed, so don’t forget to ensure that the pump is turned off during testing to avoid dead heading.
Article by John F. Higdon, PE, FASSE first appeared in ASSE’s Working Pressure magazine