A calibrated balancing valve (CBV) is, in physics terms, a fancy orifice. Once its handle is set, the flow it passes is proportional to what?

The square root of the differential pressure across it.. A set CBV follows the orifice law: flow ∝ √Δp. Double the Δp and you get about 41 % more flow; halve it and you get about 71 % of the flow. There's no spring, actuator, or compensation — it passes whatever its fixed orifice and the available Δp allow.

If system pressure changes — another zone closes, the pump curve shifts — a CBV's flow changes with it, because the valve has no compensation.

True. A CBV doesn't know what flow it's "supposed" to pass; it just obeys its orifice equation. So any Δp change from elsewhere in the system moves its flow. That's exactly why CBVs suit constant-flow (three-way) systems, where the operating point holds still, and why they're awkward on variable-flow loops where the Δp moves all day.

Why can't you balance a system of CBVs by walking the riser once and dialing each valve to its design flow?

The loads interact — throttling one branch shifts pressure to the others, changing their readings — so balancing is iterative (proportional balancing).. Because the branches share the loop, restricting the near load pushes pressure to the far loads, which then read differently. Proportional balancing picks a reference branch (often the index circuit — the longest, most-restricted), measures its flow as a fraction of design, and brings the others to the same fraction — then repeats, because each pass moves the reference. Two or three iterations is typical; it's a real day-or-more of work.

A CBV is trimmed to pass 10 GPM at 4 psi of differential pressure. With nothing else touched, the Δp across it rises to 16 psi. Roughly what flow does it pass now, in GPM?

20 GPM. Flow ∝ √Δp. The Δp went up 4× (4 → 16 psi), and √4 = 2, so flow doubles: 10 × 2 = 20 GPM. This is exactly the lack of compensation that bites on variable-flow systems — the trim that was perfect at the design Δp is wrong the moment system pressure moves.

How does an automatic balancing valve (ABV) hold design flow without a technician setting it?

A spring-loaded cartridge closes the orifice as Δp rises and opens it as Δp falls, holding flow roughly constant within its compensation range.. Inside the ABV is a spring-loaded cartridge in the flow path. Rising Δp pushes it toward closed, falling Δp lets it open, and within the cartridge's compensation range (often ~1–22 psi) the through-flow stays approximately constant. You pick the body and cartridge for the design flow, install it, and walk away — no trim, no manometer, no procedure.

An ABV that's supposed to hold a flat design flow is instead passing a flow that swings with system pressure. The cartridge isn't stuck. What's going on? expected: flat flow across the Δp range observed: flow rising/falling with system Δp cartridge: free, not jammed

The system Δp is outside the cartridge's compensation range, so it's bottomed out (fully open) or maxed out (fully closed) and behaves like a plain orifice — flow follows √Δp again.. Below the cartridge's minimum design Δp the spring can't open it any further (fully open); above the maximum it's fully compressed (as closed as it gets). At either extreme the flat-flow behaviour disappears and the valve reverts to a √Δp orifice. The fix is sizing: pick a cartridge whose compensation range comfortably brackets the Δp the system actually sees — too narrow and it spends its life outside the band where it works.

A stuck ABV cartridge is hard to diagnose because the valve has no P/T test ports — you usually need an external flow meter or added ports on the branch to confirm it.

True. That's the ABV's blind spot. Scale, debris, or age can leave the cartridge half-closed and unresponsive, and there's no way to see it from outside and no P/T ports to read Δp across the trim. Diagnosis means adding the very metering — an external flow meter or P/T ports — that an ABV was bought to avoid needing.

A pressure-independent control valve (PICV) combines which two functions in one body?

A modulating control actuator on top plus a pressure-compensating cartridge that caps flow regardless of system Δp.. A PICV stacks a control actuator (modulates 0–100 % on command) over a pressure-compensating cartridge (holds the design flow maximum regardless of system pressure). The result: the valve tells the pump exactly how much to flow and doesn't fight the changing system pressure — control and balancing in one device.

Why do PICVs pair so well with aggressive DP setpoint reset on a variable-flow system?

They hold each load's design flow across a wide Δp range, so the pump can drop its DP setpoint far without starving the worst-case load.. PICVs are pressure-independent, so each load keeps its design flow even as loop Δp swings widely. That lets the BMS push the DP setpoint down aggressively (most-open-valve reset) without far loads losing flow — which is where the cube-law pump savings live. CBVs can't do this; their flow drifts with every Δp change.

A loop that was balanced and stable starts starving its far loads right after a boiler swap — but nobody touched a balancing valve. What most likely happened?

The system curve moved (new equipment changed the resistance), so the original CBV calibration is silently invalid — the loop needs re-balancing.. A boiler swap, pump-impeller trim, system refill, or a newly tied-in load all shift the system curve. The CBVs never moved, but the operating point did, so the trim set at the old condition no longer delivers the design flows — the loop is "self-unbalanced" again with nobody having touched it. The fix is to re-walk the loop with a manometer, not to chase the symptom in the occupied space.

Practice · Hydronic Balancing

Hydronic Balancing Quiz Hydronics

Read the Hydronic Balancing lesson first, then test yourself — ten questions on the three balancing-valve families and how loops drift: CBVs and the square-root law, ABV compensation ranges, PICVs and why they pair with DP reset, and reading the symptoms of an unbalanced loop. Each reveal links back to the section it came from.

← All practice