Getting this distinction right matters for sizing, code compliance, and keeping fixtures and equipment supplied with enough pressure. Many plumbers and facility managers struggle to tell the difference between designed-in loss and a fouled or failing device.
This article breaks down what causes pressure drop, typical ranges by assembly type, how to measure it correctly, and when a drop means something's actually wrong.
Key Takeaways
- Backflow preventers drop pressure by design, using check valves and (on RPZs) a relief valve
- Typical drop at rated flow: RPZ 10-15 psi, DCVA 3-8 psi, PVB 5-10 psi (check the manufacturer curve)
- Excessive drop usually points to fouling, wrong sizing, or a failing check, not a design flaw
- Size for peak flow demand and manufacturer curves, never pipe diameter alone
What Pressure Drop Represents in a Backflow Preventer
Pressure drop is the psi differential between the supply side and discharge side, created as water passes through the assembly's check valves and, on RPZs, the relief valve mechanism. Unlike a fixed-orifice fitting with a constant Cv factor, a backflow preventer's checks open and close dynamically depending on flow rate. That means there's no single number that describes its loss. Instead, manufacturers publish a pressure-loss-versus-flow curve for each specific model and size.
Key point: pressure drop isn't something you dial up or down. It's a byproduct of the hazard-level protection the assembly is built to provide. A higher-hazard RPZ needs more internal components, so it inherently loses more pressure than a simpler double check valve assembly (DCVA).
Factors That Influence Pressure Drop in Real-World Operation
Catalog numbers reflect lab conditions at rated flow. Field losses often look different. Several factors widen that gap:
- Assembly complexity: an RPZ's added relief valve creates more restriction than a DCVA's simple two-check design
- Sizing mismatch: an oversized unit runs sluggish checks that don't seat properly; an undersized one chokes flow and spikes loss
- Fouling and wear: debris, mineral scale, and degraded rubber seats increase differential pressure over time
- Installation issues: partially closed isolation valves, undersized branch piping, and elevation changes all compound the apparent loss
Manufacturer troubleshooting guidance commonly lists a fouled strainer and an undersized valve as the two leading causes of abnormally high pressure drop. The fix in both cases is cleaning or resizing, not replacing the whole assembly.
Typical Pressure Drop Ranges by Assembly Type
Pressure loss depends on assembly design and hazard classification. Always confirm the figure against the manufacturer's flow-loss curve for your specific model rather than relying on a generic range.
Nominal Operating Range
At rated flow, on a correctly sized, new, well-maintained device:
- RPZ assemblies: typically 10-15 psi, due to the two-check-plus-relief-valve design
- DCVA assemblies: typically 3-8 psi, with a simpler two-check design
- PVB assemblies: typically 5-10 psi, used in backsiphonage-only irrigation contexts
These are reasonable planning figures. Manufacturers publish separate curves by nominal pipe size for a reason: a 3-inch and a 6-inch assembly of the same model don't lose pressure the same way at the same gpm.
Allowable Tolerance and Boundary Limits
Pressure-loss curves get steep above roughly 60% of rated capacity. That means the "rated flow" number on a spec sheet can understate real losses once demand climbs toward peak.
Distinguish between:
- Momentary spikes: irrigation cycling, multiple fixtures running at once
- Sustained undersizing: a device that's simply too small for the building's actual peak demand
The first is normal operation. The second is a sizing problem that needs correcting, not living with.
Safe Operating Margin
After you subtract every loss in the system, downstream fixtures still need enough pressure to function:
- 15-20 psi minimum for standard plumbing fixtures
- 35-50 psi for irrigation systems
Operating too close to that margin produces trickling fixtures, weak sprinkler throw, and RPZ relief valves that dribble more often than they should. Alone, none of these is a dramatic failure. Together, they signal the system has no headroom left.
How Pressure Drop Is Specified, Measured, and Validated
Pressure drop shows up in two places: on paper, as a manufacturer specification, and in the field, as something you can actually measure.
Specification and Documentation
Manufacturer flow-loss curves (not a single Cv-style number) are what you should rely on. Published "maximum allowable loss" tables from approval agencies like ASSE represent evaluation standards for product testing, not a promise of what you'll see in the field.
Measurement and Verification Methods
To measure real-world drop:
- Attach a test kit to the test cocks and record static pressure before flow begins
- Open flow to a stabilized rate and record dynamic pressure at inlet and outlet simultaneously
- Calculate the difference : that's your actual field dP at that flow rate
- Compare against the manufacturer's curve for your exact model and size
A simpler field check: compare pressure just before and just after the backflow assembly. If the drop between those two points is much larger than the curve predicts, the assembly, not the building's supply, is the likely cause.
Lab-rated curves assume clean, straight-run piping and steady lab flow. Real installations have elbows, partially open valves, and variable demand, so some deviation from the curve is normal.
Implications of Excessive or Abnormal Pressure Drop
Here's the cause-effect chain: a fouled or failing check creates turbulence, turbulence creates friction, and friction amplifies pressure loss well beyond the design curve.
Downstream symptoms to watch for:
- Low flow at fixtures and process equipment
- Sprinklers that don't throw as far as they used to
- An RPZ relief valve that's weeping or discharging more than occasionally
A continuously weeping RPZ relief valve is worth flagging clearly: the relief valve is doing its job, discharging to protect the potable water supply because the first check can no longer hold. The first check needs servicing. Never plug or disable the relief valve; that defeats the entire purpose of the assembly.
A Separate Issue: Meter Over-Registration
A different problem shows up on the metering side of a facility, separate from backflow assembly pressure loss. During demand surges, water speeds up through the meter, pressure inside the meter drops, dissolved gas comes out of solution and vapor forms, and the meter counts that volume as water. That inflates water bills even when nothing is technically "broken." Backflow assemblies can add to it: their check valves can snap shut when flow stops suddenly, sending pressure waves back toward the meter. Keeping assemblies tested and controlling water hammer at its source reduces the effect.
Water Flow Innovations' FCD addresses that problem at the meter. It creates a small, controlled back-pressure that extends back through the meter, holding pressure up during surges so far less gas and vapor form and are counted, with under 0.8 psi pressure loss across all flow conditions.
The FCD is a metering-accuracy device, not a backflow preventer. It installs immediately after the meter, ahead of the PRV and any backflow assembly (meter → FCD → PRV → backflow preventer → building distribution), and works separately from them. In that position its back-pressure reaches the meter directly. Backflow assemblies protect the supply; the FCD separately corrects what the meter records, so the meter records closer to what the facility actually receives. Learn more about correcting meter over-registration.
Frequently Asked Questions
How much pressure do you lose from a backflow preventer?
Loss depends on assembly type and flow rate: roughly 10-15 psi for RPZ assemblies, 3-8 psi for double check valve assemblies (DCVA), and 5-10 psi for pressure vacuum breakers (PVB) at rated flow. Always check the manufacturer's flow-loss curve for your specific model.
Is it normal for a backflow preventer to drip?
Occasional relief valve discharge on an RPZ during pressure fluctuations is normal. Continuous dripping means the first check valve needs servicing; the relief valve is doing its job and should never be plugged or disabled.
Can a backflow preventer cause low water pressure in a building?
Yes, by design. But an abnormally large drop usually points to fouling, undersizing, or a failing check rather than being inherent to a correctly sized, well-maintained device.
Does pipe size alone determine what size backflow preventer to install?
No. Sizing should be based on peak flow demand and available pressure, checked against the manufacturer's flow-loss curve. Oversizing causes sluggish checks; undersizing chokes flow.
How often should a backflow preventer be tested for pressure performance?
Most jurisdictions require annual testing by a certified tester, though requirements vary by state and water purveyor. Testing verifies check valve and relief valve differentials remain within specification.
What's the difference in pressure loss between an RPZ and a double check valve assembly?
RPZs typically lose more pressure than DCVAs because of the added relief valve required for high-hazard protection—often around 10-15 psi versus 3-8 psi at rated flow.
Does adding a flow conditioning device compound that pressure loss?
Only minimally. Pressure loss across the Flow Conditioning Device is under 0.8 psi across all flow conditions, unlike a backflow assembly where the drop is inherent to how it protects the supply. Downstream pressure and flow rates see no practical change.
Can demand surges in the service line reach the bill?
Yes. During demand surges, pressure inside the meter drops, gas and vapor form, and the meter counts that volume as delivered water. This happens inside the meter, separate from the backflow assembly's own designed pressure loss, though backflow check valves snapping shut can add to it by sending pressure waves back toward the meter. Documented installations typically show up to a 30% reduction in combined water and sewer costs (46% is the highest documented, a rare case).
