FLOWSYNERGY
SectionGuide
TypeArticle
Updated20 Aug 2026
StatusReference
Menu
  1. Home
  2. Guides
  3. PICV or Balancing Valve?

Guide

PICV or Balancing Valve? Choosing Flow Control for Chilled-Water Terminals

Use a pressure-independent control valve where terminal flow must hold its design value as system pressure swings, which is most variable-flow chilled-water systems with many terminals. Use a static balancing valve with a separate two-port control valve where the system is small, constant-flow or already commissioned, and where a simpler, cheaper component is easier to maintain. The deciding variables are pressure variability, commissioning effort and dirt in the loop.

A pressure-independent control valve cartridge withdrawn from its body and laid alongside it on a workshop bench.
A pressure-independent control valve cartridge withdrawn from its body and laid alongside it on a workshop bench.

What each device actually does

A pressure-independent control valve is three components in one body: a differential-pressure regulator that holds a constant drop across the control section, an adjustable flow limiter that caps maximum flow, and the control section itself, stroked by an actuator. Set the design flow on the dial and the cartridge holds it while the rest of the network moves. A static balancing valve (a double-regulating valve with a calibrated stem and a pair of test points) does one thing: it adds a fixed, measurable resistance so that, at one commissioned condition, every branch takes its share. Pair it with a separate two-port control valve and you have the older arrangement. The system-level picture is in the district-cooling and HVAC guide; the product families sit on HVAC and district-cooling valves; the head-to-head summary is PICV vs balancing valve. This article goes underneath all three.

The problem a PICV solves: terminals that take flow when their neighbours close

In a variable-flow chilled-water system, every two-port control valve that closes raises the differential pressure seen by every valve still open. A static balance is correct at exactly one set of valve positions, the one the commissioning engineer measured. Move away from it and the branches nearest the pump over-flow while the far end starves. Over-flow is not harmless: it pushes the coil past its design velocity, drops the return temperature, and pulls the network towards low-ΔT syndrome, where the plant circulates far more water than the load needs. A PICV removes the coupling. Because the internal regulator absorbs the surplus differential, terminal flow becomes a function of actuator position and dial setting, not of what the neighbours are doing. On a tower fed from a district-cooling network (the normal case in Dubai, where operators such as Empower meter energy at the substation), that independence is the difference between a network that behaves and one re-balanced at every fit-out.

The two philosophies, criterion by criterion

The table sets the trade-off out in the terms that decide it on site.

Pressure-independent control valve compared with a static balancing valve and separate two-port control valve, criterion by criterion
CriterionPICVStatic balancing valve + two-port control valve
What it controlsFlow, held to the dial settingResistance only; flow follows the ΔP imposed on it
Where it sitsOne per terminalTerminal, branch and riser, plus a control valve
ΔP rangeA stated window: below the minimum it cannot hold flow, above the maximum it is out of ratingNo window: any ΔP, but the flow moves with it
Commissioning effortSet the dial; prove riser ΔP stays in the windowFull proportional balance, measured and recorded branch by branch
Response to load changeFlow limit holds while other terminals modulateDrifts the moment the system leaves the commissioned condition
Dirt toleranceLow: fine clearances and a small orifice at low flowsHigh: an open waterway with a measuring orifice
Cost bandHigher per unit, fewer commissioning hoursLower per unit plus a control valve, more commissioning hours
When it winsVariable flow, many terminals, phased fit-out, high riser ΔPSmall or constant-flow systems, documented existing balance, thin maintenance capability
Decision table for terminal flow control. Ranges and cost bands are indicative; the ΔP window and flow-setting chart for the specific model govern.

Static balancing done properly is not the weak option

A proportionally balanced system, measured with a calibrated differential-pressure meter and recorded branch by branch, is a legitimate engineering answer. The discipline is codified in CIBSE Commissioning Code W and in the hydronic design guidance published by ASHRAE, both unpacked in the glossary (CIBSE, ASHRAE). It wins where flow does not vary much: constant-flow primary loops, small plantrooms, retrofits where the existing balance is documented and still valid. It also wins where maintenance capability is thin. A double-regulating valve has an open waterway, a visible stem and a setting you can read off the handwheel, and it can be stripped and put back by anyone competent with a spanner. Where it loses is the phased tower: a floor commissioned in year one is quietly un-balanced by the floor connected in year three, and nobody returns with the meter.

Mechanical versus electronic: the cartridge does the work, the actuator only modulates

A mechanical PICV does its pressure work with a spring and a diaphragm, and needs no power to do it; the actuator on top strokes the control section and nothing else. An electronic PICV replaces the mechanical regulator with a measured-flow control loop (a flow sensor, a controller and a bus connection) and buys real things with it: live flow and energy data, remote re-setting, a diagnostic trail. Our position rests on failure modes, not features. When a mechanical cartridge loses its actuator it still limits flow: the coil is throttled, the network is not destabilised. When an electronic valve loses its sensor, its bus address or its commissioning file, it loses the flow limit itself, and the fault presents as a comfort complaint three floors away.

The cartridge already solved the hard problem. Buy the electronics for the data and the remote reset when the project will genuinely use them, not to fix a balance the mechanical regulator has already fixed, and never on a site with nobody to own the configuration after handover.

The position this guide takes: editorial, not a standard

Dirt: why cartridges foul, and the strainer rule

The regulating element controls with small clearances and, at low design flows, a small orifice. Magnetite from an unflushed system, jointing compound, weld scale and construction debris all arrive there, and the failure is undramatic: the valve simply stops holding its setting, usually on the terminals with the lowest flow first. Two rules follow. A strainer belongs at every terminal, not only at the riser. A riser strainer protects the riser. And the loop is flushed and chemically cleaned before the cartridges are commissioned, with flushing bypasses that exist in steel rather than on a drawing. Where nobody can vouch for the loop being clean (an existing system being extended, a fast-track tower handed over in phases), the honest answer is often a serviceable balancing arrangement now and pressure independence at the next refurbishment.

Authority, Kv and the ΔP window: reading the flow chart correctly

A conventional two-port valve only controls if it owns a decent share of the branch resistance. Valve authority is the ratio of the drop across the fully open valve to the drop across the whole controlled circuit, and it is commonly kept above about half for a modulating coil. That calculation is why a control valve is chosen on Kv rather than on pipe size, and why the line-size valve is almost always the wrong one. A PICV changes the question rather than removing it: authority is effectively fixed by the internal regulator, so what you read from the chart instead is the ΔP window: the minimum differential below which the regulator cannot maintain set flow, and the maximum above which it is outside its rating. Size riser and branch so every terminal sits inside that window at both extremes: full design flow with everything open, and minimum flow in the small hours.

Reference values. Confirm against the current edition of the standard and the manufacturer's rating table.

Worked example: a fan-coil floor on a district-cooling riser

Take a floor of twenty-four fan-coil units, each with a design flow of 0.25 l/s, on a riser whose differential pressure swings between 40 and 180 kPa as the rest of the tower modulates. Sized the static way, each terminal takes a two-port control valve chosen for authority at the design condition plus a double-regulating valve set during a proportional balance: several days of commissioning for the floor, and a result correct on the day it was measured. Sized with PICVs, each terminal takes one valve, the dial is set to the design flow, and the commissioning task becomes proving the riser differential stays inside the cartridge window across the load range. The materials bill is higher; the commissioning bill is lower, and the floor still holds its flows when the storeys above are fitted out.

Reference values. Confirm against the current edition of the standard and the manufacturer's rating table.

Cost over the life of the system, and the two buyer mistakes

Both mistakes cost the same money at different times. Too cheap buys downtime: the lowest-priced cartridge dropped into a loop nobody flushed, with no terminal strainers, fouls within a season, and the cost surfaces as comfort complaints, re-commissioning and valves replaced behind ceilings. Over-engineered buys cost: electronic valves with a bus drop on every coil of a small constant-flow system, where a commissioned double-regulating valve would have held flow for two decades and the data will never be read. The honest comparison is not unit price against unit price. It is unit price plus commissioning hours plus the re-balance you will or will not have to fund when the next tenancy connects.

Governing documents and the confirm-the-edition rule

Terminal valve bodies and their ends are rated under the same documents as any other valve: ASME B16.34 for pressure-temperature ratings, ASME B16.5 for flanged ends, ISO 5208 or API 598 for shell and seat testing, EN 10204 for the inspection documents that prove the material. Balancing and PICV performance itself is not fixed by a single public standard: flow-setting tables, ΔP windows and accuracy classes belong to the manufacturer and differ between models in one catalogue. Treat every figure here as reference data. Selecting by service condition runs the same sequence for the rest of the network, and what the Gulf changes about district-cooling valve selection covers the climate-specific part.

Reference values. Confirm against the current edition of the standard and the manufacturer's rating table.

Documentation to request, and the next step

Ask for four things with the quotation, whichever arrangement you choose: the flow-setting table or chart for the exact model, with dial positions; the stated ΔP window, minimum and maximum; the body pressure-temperature rating and end-connection standard; and the material and inspection documents for the pressure-retaining parts. For a static arrangement, add the Kv table and the calibrated stem-position chart. Further guides sit on Insights.

Specify the duty first. Send the medium, pressure, temperature, size and quantity. View our trading brands or ask our Dubai team to confirm a suitable product.

Enquiry Dubai

Tell us the service condition.

Medium, operating and design pressure, operating and design temperature, size and end connection, and quantity. An enquiry that carries the full service condition is answered as a selection against the governing standard. Price, lead time and availability are confirmed per enquiry, never assumed.

Send an enquiry
MediumChilled water, steam, seawater, gas, process fluid
PressureOperating and design, bar g
TemperatureOperating and design, °C
Size & connectionDN / NPS, flanged, threaded or welded