How to Design a Chilled Water Piping System Step by Step
- nexoradesign.net
- 5 days ago
- 24 min read
Designing a chilled water piping system is not simply a matter of selecting a pipe diameter from a chart. A complete design must connect the building cooling load, chilled-water temperature difference, terminal-unit flow rates, system arrangement, pipe pressure losses, pump duty, valve authority, chiller minimum-flow limits, pressurization, controls and commissioning requirements.
If any one of those decisions is made in isolation, the finished system may suffer from low delta-T, noisy valves, excessive pumping energy, poor coil control, insufficient chiller flow, condensation, trapped air or chronic balancing problems.
This guide explains how to design a chilled water piping system step by step, using SI units first and practical engineering checks throughout. A worked example follows a 1,200 kW system from cooling load to preliminary pump duty.
Important: This article is an engineering workflow, not a substitute for the adopted code, authority requirements, project specifications or current manufacturer data. Final selections must be checked by the responsible engineer.
New to HVAC design? Build the wider foundation—chillers, pumps, piping, valves, controls, commissioning and the complete design workflow—with the HVAC Beginner’s Guide.
Chilled Water Piping Design Workflow at a Glance
Step | Design task | Main output |
1 | Establish the design basis | Loads, temperatures, fluid, redundancy and codes |
2 | Select CHWS, CHWR and design delta-T | Agreed water-temperature regime |
3 | Calculate terminal and total water flow | Coil-flow and plant-flow schedule |
4 | Select the hydraulic arrangement | Constant primary, primary-secondary, variable primary or district interface |
5 | Develop the schematic and pressure zones | Design schematic and zoning strategy |
6 | Route the piping and assign segment flows | Pipe network with flow in every section |
7 | Select pipe material and pressure class | Material, schedule, joints and pressure rating |
8 | Size every pipe section | Diameter, velocity and friction schedule |
9 | Calculate pipe, fitting and equipment losses | Pressure-loss schedule |
10 | Identify the critical hydraulic circuit | Index circuit for pump-head calculation |
11 | Select and verify the pumps | Flow, head, curve, power, NPSH and redundancy |
12 | Select valves and accessories | Coil and plant valve schedules |
13 | Design pressurization and air/dirt removal | Expansion vessel, fill pressure, separators, vents and drains |
14 | Design insulation, supports and thermal movement | Insulation and pipe-support details |
15 | Define controls and minimum-flow protection | BMS sequence and control points |
16 | Plan TAB and commissioning | Test, balance, trend and handover requirements |
Main Components of a Chilled Water System
A typical closed-loop chilled-water system may include:
Air-cooled or water-cooled chillers
Primary, secondary or variable-primary chilled-water pumps
Chilled-water supply and return headers
AHU, FAHU and FCU cooling coils
Two-way control valves, PICVs or balancing valves
Isolation and check valves
Strainers, dirt separators and air separators
Expansion vessel and pressurization unit
Flow meters, temperature sensors, pressure gauges and P/T ports
Automatic air vents at high points and drains at low points
Differential-pressure sensors and pump VFDs
Bypass piping when required for minimum chiller flow
Pipe insulation with a continuous vapour barrier
Water-treatment, flushing and chemical-dosing provisions
The components must be designed as one hydraulic system. For example, a control valve cannot be selected correctly until its flow and available differential pressure are known; the pump cannot be selected correctly until the critical circuit is known; and a variable-primary system cannot be finalized until the selected chiller’s allowable flow range and rate of flow change are confirmed.
Step 1 — Establish the Design Basis (How to Design a Chilled Water Piping System Step by Step)
Do not begin pipe sizing until the design basis is written down and agreed. At minimum, record the following information.
Design input | Questions to resolve |
Cooling loads | Are the values room peaks, block peaks, coil loads or plant loads? Has diversity been applied correctly? |
Design conditions | Which indoor and outdoor conditions, schedules and ventilation rates were used? |
CHWS and CHWR | What temperatures are required at the chiller and at the coils? |
Fluid | Treated water or water-glycol? What concentration and design temperature? |
Operating profile | Is the load stable, highly variable, 24/7 or seasonal? |
Redundancy | N, N+1, duty/standby or no standby? Which loads are essential? |
Future capacity | Is a genuine future load identified, or is extra capacity only an undefined allowance? |
Building height | Will one pressure zone exceed the rating of lower-level equipment? |
Plant arrangement | Air-cooled chiller, water-cooled chiller, heat exchanger or district-cooling interface? |
Codes and standards | Which local code, authority standard, pipe code and energy requirements apply? |
Equipment data | What are the selected chiller, coil, valve and strainer pressure drops and flow limits? |
Start with the correct load boundary
The sum of individual room peak loads is not automatically the coincident building peak. Similarly, room sensible load is not the same as cooling-coil duty. Ventilation, latent load, fan heat, duct heat gain, system diversity and operating schedules can change the load seen by the coil and plant. (How to Design a Chilled Water Piping System Step by Step)
For piping design, use the selected coil duty and design water temperature difference for each terminal unit. Use the coincident plant duty for the main plant flow. Keep those boundaries clear so diversity is not applied twice.
Need a defensible cooling-load basis before calculating water flow? The Cooling Load Design Handbook covers room, zone, system, building and coil-load boundaries, psychrometrics, airflow, coil duty and equipment-selection strategy.
Step 2 — Select Chilled-Water Supply, Return and Delta-T
Common chilled-water temperature regimes include 6/12°C, 7/12°C and project-specific higher-delta-T selections. None is universal. The correct temperatures must be coordinated with:
Chiller performance and allowable leaving-water temperature
Coil capacity and row depth
Required supply-air temperature and dehumidification
Control-valve selection
Pumping energy and pipe sizes
District-cooling requirements, where applicable
Risk of condensation or freezing
Water-glycol concentration and properties
The design temperature difference is:
ΔT = CHWR temperature − CHWS temperature
A larger delta-T reduces design water flow for the same cooling duty. That can reduce pipe size and pumping energy, but the coils and chillers must be selected to deliver the required duty at that temperature regime. A high design delta-T on paper is useless if the coil selection, controls and operating conditions cannot maintain it.
Effect of delta-T on flow for a 1,200 kW load
CHWS/CHWR | ΔT | Required water flow |
7/12°C | 5 K | 57.3 L/s or 206.3 m³/h |
6/12°C | 6 K | 47.8 L/s or 172.0 m³/h |
6/14°C | 8 K | 35.8 L/s or 129.0 m³/h |
The table shows why delta-T matters, but it does not prove that 8 K is suitable. Confirm the actual coil and chiller selections.
Step 3 — Calculate Chilled-Water Flow Rate
The fundamental heat-balance equation is:
Q̇ = ṁ × cₚ × ΔT
Because mass flow is related to volume flow:
Q̇ = ρ × V̇ × cₚ × ΔT
Where:
Q̇ = cooling duty, kW
ṁ = mass flow rate, kg/s
V̇ = volume flow rate, m³/s
ρ = fluid density, kg/m³
cₚ = specific heat, kJ/(kg·K)
ΔT = water temperature difference, K
For water at typical chilled-water temperatures, practical SI equations are:
V̇ (L/s) = Q̇ (kW) ÷ [4.186 × ΔT (K)]
V̇ (m³/h) = Q̇ (kW) ÷ [1.163 × ΔT (K)]
For IP units:
Flow (GPM) ≈ Cooling load (Btu/h) ÷ [500 × ΔT (°F)]
Worked flow calculation
Assume:
Coincident cooling duty = 1,200 kW
CHWS = 6°C
CHWR = 12°C
ΔT = 6 K
Fluid = treated water
Then:
V̇ = 1,200 ÷ (4.186 × 6) = 47.8 L/s
V̇ = 47.8 × 3.6 = 172.0 m³/h
Terminal-unit flow schedule
Load group | Coil duty | Design flow at 6 K |
AHU-1 | 300 kW | 11.94 L/s |
AHU-2 | 240 kW | 9.56 L/s |
AHU-3 | 210 kW | 8.36 L/s |
FCU Group A | 250 kW | 9.95 L/s |
FCU Group B | 200 kW | 7.96 L/s |
Total | 1,200 kW | 47.78 L/s |
The sum of branch flows must reconcile with the plant flow. If the selected coils use different water-temperature differences, calculate each coil from its own selected duty and flow rather than forcing one generic factor onto the whole schedule.
Water-glycol warning
Do not use the water constants unchanged for glycol. Glycol changes density, specific heat and viscosity. It can require more flow, increase pressure drop and change pump power and NPSH. Use the fluid manufacturer’s properties at the actual mean operating temperature and concentration.
Step 4 — Choose the Piping and Pumping Configuration
The system arrangement determines how water flow is controlled and how chillers are protected.
Option A: Constant-primary-flow system
A constant-flow pump circulates water through the chiller and distribution system. Three-way valves or constant-flow branches may maintain circulation as loads change.
Advantages: simple control and predictable chiller flow.
Limitations: unnecessary bypass flow and higher pumping energy at part load.
Best fit: small or stable-load systems where simplicity is more important than part-load optimization.
Option B: Primary-secondary system
Primary pumps circulate water through the operating chillers. Secondary VFD pumps serve the distribution system. A low-loss common pipe hydraulically decouples the two circuits.
Advantages: predictable chiller flow, robust staging and clear hydraulic separation.
Limitations: more pumps, more space and possible supply/return mixing at the common pipe.
Best fit: multi-chiller plants requiring conservative, well-understood operation.
Option C: Variable-primary-flow system
One VFD pump set moves water through both the chillers and the distribution system. Two-way valves reduce flow as the cooling load falls.
Advantages: fewer pumps and potentially lower pumping energy.
Limitations: chiller minimum flow, maximum flow, flow measurement, bypass control and staging transitions must be carefully engineered.
Best fit: variable-load plants with compatible chillers, reliable controls and a fully developed operating sequence.
For variable-primary systems, verify the selected chiller’s permitted evaporator flow range and rate of flow change at every staging condition. Bell & Gossett’s Variable Primary Flow Systems manual explains why minimum flow and staging transitions must be designed—not assumed.
Option D: District-cooling or heat-exchanger interface
The building side is hydraulically separated from the district side by a plate heat exchanger or energy-transfer station. The designer must coordinate:
Contracted supply/return temperatures and required delta-T
Maximum primary-side flow
Heat-exchanger approach temperature
Primary and secondary pressure drops
Control-valve authority
Metering requirements
Building-side pressure zone and expansion system
Configuration comparison
Arrangement | Pumping concept | Part-load efficiency | Control complexity | Main design risk |
Constant primary | One constant-flow circuit | Low to moderate | Low | Excess bypass flow and energy |
Primary-secondary | Constant primary + variable secondary | Good | Moderate | Common-pipe mixing and excess pumps |
Variable primary | One variable-flow circuit | Potentially high | High | Chiller minimum flow and staging |
District interface | Primary utility side + building secondary side | Project-dependent | Moderate to high | Contract delta-T and interface control |
Direct return vs reverse return
In a direct-return system, the first coil supplied is also the first coil returned. Circuit lengths are unequal, so deliberate balancing is required.
In a reverse-return system, the first coil supplied is the last coil returned. Total pipe lengths are more nearly equal, which can reduce the balancing burden when terminal units and branch losses are similar.
Item | Direct return | Reverse return |
Pipe quantity | Lower | Higher |
Natural hydraulic balance | Lower | Better when branches are similar |
Balancing requirement | Essential | Still required and verified |
Flexibility | High | Best with repeated similar terminals |
Typical use | Compact systems with PICVs or balancing valves | Long repeated branches, risers or similar coils |
Reverse return does not guarantee equal flow when coil, valve and branch pressure drops differ. Bell & Gossett’s Hydronic System Types manual provides useful system-arrangement fundamentals.
Step 5 — Develop the Schematic and Pressure Zones
Before detailed routing, prepare a clear system schematic showing:
Chillers, pumps and heat exchangers
CHWS and CHWR flow directions
Equipment isolation valves
Check valves and pump discharge arrangements
Coil control and balancing valves
Strainers and dirt separators
Expansion-vessel connection
Air-separator, vent and drain locations
Flow meters and temperature/pressure sensors
DP sensor locations
Minimum-flow bypass, if required
Normal, standby and future equipment
Pressure zones and design pressures
High-rise static-pressure check
The static pressure at the lowest point is approximately:
p = ρ × g × h
For water, every 10 m of elevation is approximately 98.1 kPa, or 0.981 bar. A 100 m water column therefore produces about 981 kPa, or 9.81 bar, before adding fill-pressure margin and pump differential pressure.
Compare the maximum operating and test pressures with the rating of every chiller evaporator, coil, valve, pipe, flexible connector and accessory. Tall buildings may require separate pressure zones, pressure-break heat exchangers or specially rated equipment.
Remember the distinction:
Static height affects fill pressure, expansion-vessel precharge and component pressure rating.
Static height does not get added to steady-state circulating pump head in a sealed closed loop.
For a deeper treatment, see Nexora’s guide to chilled-water system design for high-rise buildings.
Step 6 — Route the System and Allocate Flow to Every Pipe Segment
Develop the routing on coordinated floor plans and riser diagrams. Then assign a design flow to every segment by adding the connected downstream coil flows.
Using the worked example, the supply-main flow reduces after each take-off:
Main segment | Connected downstream load | Segment flow |
Plant to AHU-1 take-off | 1,200 kW | 47.78 L/s |
After AHU-1 | 900 kW | 35.83 L/s |
After AHU-2 | 660 kW | 26.28 L/s |
After AHU-3 | 450 kW | 17.92 L/s |
After FCU Group A | 200 kW | 7.96 L/s |
The return-side flow builds in the opposite direction. Mark each segment with a unique reference so the pipe-sizing and pressure-loss schedules can be audited against the drawings.
When routing, also coordinate:
Ceiling and riser space
Access to valves and strainers
Drainage from low points
Venting at high points
Pipe slope where the specification requires it
Insulation thickness and vapour-barrier clearance
Structural openings and sleeves
Expansion movement, anchors and guides
Equipment-removal paths
Electrical, fire-protection and architectural services
Step 7 — Select Pipe Material, Schedule and Pressure Class
Pipe selection depends on diameter, pressure, temperature, water quality, fire rating, joining method, local availability and the project specification.
Material | Common application | Main checks |
Carbon steel | Plant rooms, risers and large mains | Corrosion allowance, water treatment, welding/grooving, pressure class |
Copper | Small equipment branches | Cost, jointing, velocity, erosion/corrosion and dissimilar metals |
Stainless steel | Corrosion-sensitive or high-quality systems | Grade, jointing, cost and compatibility |
PP-R or approved plastic systems | Selected branches and distribution | Temperature/pressure derating, oxygen diffusion, high thermal expansion, supports and fire requirements |
Pre-insulated steel or HDPE systems | Buried or district-energy piping | Burial design, joints, leak detection, expansion and manufacturer system limits |
Use the actual internal diameter, not only the nominal diameter, when calculating velocity and friction. The internal diameter changes with material, wall thickness and pipe schedule.
For building-services piping, ASME B31.9 is one relevant reference where adopted. The governing local code and project specification must determine the final design, fabrication, inspection and testing requirements.
Step 8 — Size the Chilled-Water Pipes
Pipe sizing is an optimization between:
Acceptable water velocity
Acceptable friction loss per metre
Pumping energy
First cost and available space
Noise and vibration
Erosion and water hammer risk
Air transport and air-separation strategy
Control-valve differential pressure
Future and standby operating scenarios
Velocity equation
v = 4V̇ ÷ (π × Dᵢ²)
Where:
v = velocity, m/s
V̇ = volume flow, m³/s
Dᵢ = actual pipe internal diameter, m
Preliminary screening ranges
The following values are reasonable starting points for preliminary screening, not universal code limits. The project specification, pipe material, acoustic criteria, water treatment and selected equipment may require different limits.
Pipe service | Preliminary velocity check | Preliminary friction check |
Small terminal branches | About 0.6–1.2 m/s | About 100–300 Pa/m |
Distribution branches and risers | About 0.9–1.8 m/s | About 100–350 Pa/m |
Large mains | About 1.2–2.4 m/s | About 80–300 Pa/m |
In a variable-flow system, velocity will naturally fall below the design-flow screening range at part load. Provide effective central air separation and high-point venting instead of relying on water velocity alone to move air.
Bell & Gossett’s Hydronic System Design with System Syzer discusses preliminary velocity and friction considerations. Its cited limits are design guidance, not a replacement for project criteria.
Candidate-size check for the 47.78 L/s main
Assume, for illustration only:
Candidate A internal diameter = 202.7 mm
Candidate B internal diameter = 154.1 mm
Water density = 998 kg/m³
Preliminary Darcy friction factor = 0.020
Candidate | Velocity | Approximate straight-pipe friction | Comment |
202.7 mm ID | 1.48 m/s | 108 Pa/m | Reasonable preliminary candidate |
154.1 mm ID | 2.56 m/s | 425 Pa/m | High velocity and friction for many comfort projects |
The smaller pipe may cost less initially but increases pressure loss and pump energy for the life of the building. Complete a life-cycle and space check rather than selecting only by capital cost.
Read the detailed Nexora guide on sizing chilled-water pipes using velocity and friction limits for a deeper explanation.
Step 9 — Calculate Pipe, Fitting and Equipment Pressure Losses
Straight-pipe loss: Darcy-Weisbach equation
Δpₚᵢₚₑ = f × (L ÷ Dᵢ) × (ρv² ÷ 2)
Where:
f = Darcy friction factor
L = pipe length, m
Dᵢ = internal diameter, m
ρ = fluid density, kg/m³
v = velocity, m/s
The friction factor depends on Reynolds number and relative roughness. Use a recognized equation or hydraulic calculation tool, with roughness appropriate to the pipe material and condition.
Fitting loss: K-factor method
Δp𝒻ᵢₜₜᵢₙ𝓰ₛ = ΣK × (ρv² ÷ 2)
Include elbows, tees, reducers, enlargers, branch connections and other fittings on the actual circuit. A valve or accessory may be represented by a manufacturer pressure drop or by a K-factor when suitable published loss data is unavailable. Alternatively, use a consistent equivalent-length method. Do not mix methods in a way that counts the same item twice.
Equipment losses
Obtain design-flow pressure drops from the selected manufacturers for:
Chiller evaporators
Plate heat exchangers
AHU and FCU coils
Control valves and PICVs, recorded as the separate control-valve term in the pump-head calculation below
Balancing valves
Strainers and dirt separators
Flow meters
Check valves
Plant isolation valves and specialty items
Equipment pressure drop normally changes approximately with the square of flow in the turbulent region, but use manufacturer data for final calculations—especially for valves and complex equipment.
Count every loss once. In this article, the control valve or PICV is kept as a separate Δp control term and is excluded from the equipment subtotal. A check valve with a published design-flow pressure drop is not also entered as a K-factor fitting.
Pressure-loss schedule format
Circuit item | Flow | Size/model | Quantity/length | K or pressure drop | Total pressure drop |
Supply pipe | L/s | DN / ID | m | Pa/m | kPa |
Supply fittings | L/s | Type | No. | K | kPa |
Coil | L/s | Selected model | 1 | Manufacturer data | kPa |
Control valve/PICV | L/s | Selected model | 1 | Manufacturer data | kPa |
Return pipe and fittings | L/s | DN / ID | m / No. | Pa/m or K | kPa |
Chiller evaporator | L/s | Selected model | 1 | Manufacturer data | kPa |
Step 10 — Identify the Hydraulically Critical Circuit
The critical circuit—also called the index circuit—is the complete closed path with the greatest total resistance at its design flow. It is not necessarily the geographically farthest coil.
Trace from the pump through:
The common supply piping
The relevant branch and riser
The terminal-unit coil
The control or balancing valve
The return branch, riser and main
Any chiller evaporator or heat exchanger that is hydraulically in series with that pump
All common plant fittings and accessories
Calculate several likely circuits if the answer is not obvious.
Count only losses in the hydraulic circuit served by the pump being selected. For example, a secondary distribution pump in a primary-secondary system normally excludes the chiller evaporator and primary loop; the primary pump is calculated separately through its own chiller circuit. In the variable-primary worked example below, the evaporator is in series with the distribution pump and is therefore included.
Closed-loop static-head rule
In a sealed closed loop, the upward static head in the supply is balanced by the downward static head in the return. Therefore, building height is not added to the operating pump head.
Building height still affects fill pressure, expansion-vessel precharge, top-of-system pressure, NPSH and component pressure rating. Open systems must include any uncompensated static lift.
Pump-head equations
Δpₚᵤₘₚ = ΣΔpₚᵢₚₑ + ΣΔp𝒻ᵢₜₜᵢₙ𝓰ₛ + ΣΔpₑᵩᵤᵢₚₘₑₙₜ + Δp𝚌ₒₙₜᵣₒₗ
Hₚᵤₘₚ = Δpₚᵤₘₚ ÷ (ρ × g)
For water, 100 kPa is approximately 10.2 m of water head.
Worked critical-circuit loss
Item | Pressure drop |
Supply piping | 86 kPa |
Return piping | 74 kPa |
Pipe fittings | 24 kPa |
Chiller evaporator | 60 kPa |
Remote AHU coil | 45 kPa |
PICV/control valve | 30 kPa |
Strainer and flow measurement | 15 kPa |
Common plant valves/accessories | 18 kPa |
Calculated total | 352 kPa |
The calculated head is:
H = 352,000 ÷ (998 × 9.81) = 35.9 m
If a documented preliminary-design allowance of 5% is approved to cover remaining minor uncertainty:
Preliminary pump head = 35.9 × 1.05 = 37.7 m, rounded for selection to approximately 38 m.
Do not add a large arbitrary margin to compensate for missing calculations. Oversized pumps create excess flow, valve noise, unstable control and wasted energy.
Want the loss schedule and pump duty organized in Excel? Use the Nexora Pump Head Calculator Design Suite v1.0 to calculate pipe, fitting and equipment losses, pump head, pressure, motor-power estimate and NPSH check, with a printable engineering report.
Step 11 — Select and Verify the Chilled-Water Pump
The preliminary duty from the worked example is:
Flow = 47.8 L/s
Head = approximately 38 m
Do not stop at those two numbers. Check the following items on the selected pump curve:
1. Duty point and best-efficiency region
The design point should lie within the manufacturer’s acceptable operating region, reasonably near the pump’s efficient zone—not at the extreme end of the curve.
2. Complete operating range
For a VFD pump, plot or assess the system curve and all expected load conditions. Verify stable operation at minimum flow, staging transitions and duty/standby changeover.
3. Shut-off head and component rating
Check the maximum pressure that can occur at pump shut-off or at maximum VFD speed. Confirm that the pipework, valves, chiller and coils remain within their pressure ratings.
4. NPSH
Confirm:
NPSH available > NPSH required + manufacturer/project margin
Evaluate the worst credible fluid temperature, fill pressure, strainer condition and plant configuration. The expansion-vessel connection and pump suction pressure affect NPSH available.
5. Motor and drive
Hydraulic power is:
Pₕ = ρ × g × V̇ × H
Estimated electrical input is:
Pᵢₙ = (ρ × g × V̇ × H) ÷ (ηₚ × ηₘ × ηᵥ𝒻𝒹)
Using the example:
ρ = 998 kg/m³
V̇ = 0.0478 m³/s
H = 38 m
Pump efficiency = 78%
Motor efficiency = 93%
VFD efficiency = 97%
Estimated input power ≈ 25.3 kW at the stated duty.
The final motor rating must be checked against the full selected curve, service factor, fluid density, maximum speed and manufacturer data. Do not select a motor from the single calculated number alone.
6. Redundancy and parallel operation
Define whether the system uses 1 duty + 1 standby, 2 duty + 1 standby, or another arrangement. For parallel pumps, verify the combined curve and ensure each operating combination remains stable and efficient.
Calculate your own project duty: Open the Nexora Pump Head Calculator Design Suite v1.0.
Step 12 — Select Valves and Coil Accessories
A coil branch must be controllable, measurable, isolatable, ventable, drainable and serviceable.
Figure 4 illustrates a PICV providing both two-way modulation and flow limitation. If a conventional pressure-dependent two-way control valve is used instead, add the balancing or differential-pressure-control arrangement required by the hydraulic design.
Component | Purpose | Main selection check |
Isolation valve | Isolate the coil or equipment | Pressure class, full-port requirement, access |
Y-strainer/dirt protection | Protect coil and valve | Mesh, pressure drop, blowdown and cleaning access |
P/T ports or gauges | Verify temperature and pressure | Accessibility and sensor range |
Two-way control valve | Modulate water flow | Cv/Kvs, authority, close-off pressure, leakage and actuator |
PICV | Combine control and pressure-independent flow limitation | Design flow, minimum/maximum DP range, close-off and commissioning access |
Manual/automatic balancing valve | Establish design flow | Flow range, available DP and measurement method |
Check valve | Prevent reverse flow | Type, cracking pressure, noise and pressure drop |
Automatic air vent | Release trapped air at local high point | Isolation, leakage control and access |
Drain valve | Drain coil/branch at low point | Hose connection, access and safe discharge |
Control-valve authority
For a pressure-dependent control valve:
Valve authority, a = Δpᵥ ÷ (Δpᵥ + Δpᵣₑₛₜ)
Where Δpᵥ is the pressure drop across the fully open valve at design flow, and Δpᵣₑₛₜ is the pressure drop across the remainder of the controlled circuit at the same flow.
Very low authority produces unstable or non-linear control. Select the valve from the actual branch pressure conditions, required flow, allowable pressure drop and actuator close-off capability.
A PICV can simplify hydraulic control by maintaining a pressure-independent relationship between actuator position and flow, while limiting the preset maximum flow, but only within its published differential-pressure range. TAB must verify the maximum-flow setting and the available differential pressure.
Step 13 — Design Expansion, Pressurization and Air/Dirt Removal
Expansion vessel
The expansion vessel maintains acceptable system pressure as fluid temperature and volume change. Determine:
Total system fluid volume
Fluid type and glycol concentration
Fill temperature and maximum operating temperature
Static height above the vessel connection
Minimum pressure required at the highest point
Maximum allowable pressure of the weakest component
Vessel precharge
Acceptance fraction and reserve volume
The fluid-volume change may be estimated from specific volumes:
ΔV = Vₛᵧₛ × [(v₂ ÷ v₁) − 1]
The required vessel size is then related to the usable acceptance fraction between the minimum and maximum pressures. Use a recognized sizing method and the selected vessel manufacturer’s data.
Connect the expansion vessel at the point of no pressure change, normally close to the pump suction. Keep the connection hydraulically open during normal operation; an accidentally closed isolation valve can remove the vessel from the system.
Air and dirt separation
Air separates most readily where water temperature is higher and pressure is lower. In many chilled-water systems, the return near the pump suction is a suitable location for the central air separator. Caleffi explains the principle in its guidance on air-separator location.
Also provide:
Automatic air vents at local high points
Manual vents where automatic vents are unsuitable
Drain valves at low points
Dirt or magnetic separation where required
Side-stream filtration or water-treatment provisions where specified
Adequate access for cleaning and maintenance
For a detailed calculation, see Nexora’s guide on sizing expansion tanks for chilled-water systems.
Step 14 — Design Insulation, Supports and Thermal Movement
Condensation-control insulation
Chilled-water insulation must keep the outer surface above the surrounding dew point and prevent moisture from reaching the cold pipe.
Specify:
Insulation material and thickness
Thermal conductivity at the relevant mean temperature
Vapour-permeance requirement
Continuous sealed vapour barrier
Insulated pipe supports or load-bearing inserts
Sealed joints at valves, flanges, hangers and penetrations
Outdoor weatherproof cladding and UV protection
Fire and smoke performance required by code
A small gap or puncture in the vapour barrier can create concealed condensation even when the nominal insulation thickness is adequate.
Thermal movement
Pipe movement is estimated by:
ΔL = α × L × ΔT
Where:
ΔL = change in length
α = coefficient of linear expansion
L = pipe length
ΔT = temperature change from installation condition
Coordinate anchors, guides, loops, offsets and approved flexible elements. Do not use flexible connectors as a general substitute for engineered pipe flexibility.
Supports
Check support spacing, concentrated loads at valves, seismic/wind restraint where required, vibration isolation and access. The support detail must preserve the insulation and vapour barrier without compressing them.
Step 15 — Define the Control Sequence
Controls are part of the hydraulic design, not an afterthought.
Typical variable-flow sequence
Condition | Control action |
Cooling demand starts | Enable lead pump, prove flow, then enable chiller according to manufacturer sequence |
Coil load changes | Two-way valves modulate to maintain leaving-air or space-temperature setpoint |
DP changes | Pump VFD maintains the required differential pressure at the critical area |
Low load | Reduce pump speed while preserving required valve differential and chiller flow |
Flow approaches chiller minimum | Open controlled minimum-flow bypass or stage equipment as designed |
Additional capacity required | Open/isolate chiller path in the correct sequence and stage the next chiller |
Load falls | De-stage without exceeding the permitted evaporator flow-change rate |
No cooling demand | Disable chiller and pump after the specified run-on period |
Locate the main DP sensor near the hydraulically critical coil-and-valve branch, not automatically at the physically farthest point. Final position must be coordinated with the system topology and control strategy.
The U.S. Department of Energy’s School Chiller Replacement Package describes remote DP sensing, VFD control, minimum-flow protection and valve-position-based reset concepts. Avoid competing reset loops: CHWS reset and DP reset must be coordinated so the system does not respond ambiguously to the same valve-position signal.
Pump affinity laws
For the same pump and impeller, approximately:
Flow ∝ speed
Head ∝ speed²
Power ∝ speed³
These relationships explain why variable-speed pumping can reduce part-load energy, but actual system performance depends on constant control-head requirements, any true static lift in open systems, minimum speed, pump efficiency and controls.
Complete Worked Example Summary
Item | Selected/calculated value |
Coincident cooling load | 1,200 kW |
CHWS / CHWR | 6°C / 12°C |
Design delta-T | 6 K |
Fluid | Treated water |
Total design flow | 47.78 L/s or 172.0 m³/h |
Hydraulic arrangement | Variable primary flow with two-way valves/PICVs |
Distribution | Direct return with engineered balancing |
Main candidate | Approx. 202.7 mm internal diameter |
Main velocity | Approx. 1.48 m/s |
Main friction | Approx. 108 Pa/m using stated assumptions |
Critical-circuit pressure loss | 352 kPa calculated |
Preliminary design allowance | 5%, documented for remaining uncertainty |
Preliminary pump duty | 47.8 L/s at approximately 38 m |
Estimated input power at duty | Approx. 25.3 kW using stated efficiencies |
This example is preliminary. Final design requires actual pipe schedules, fitting quantities, selected equipment losses, valve data, pump curves, NPSH, chiller flow limits and the approved control sequence.
Move from a list of losses to a professional calculation report: Use the Nexora Pump Head Calculator Design Suite v1.0.
Required Drawings, Schedules and Specifications
A complete chilled-water piping design package should normally include:
Deliverable | Minimum content |
Design schematic | Equipment, flow direction, valves, sensors, bypasses, expansion and pressure zones |
Floor plans | Pipe routing, sizes, elevations, slopes, valves and equipment connections |
Riser diagram | Vertical distribution, zones, sizes, vents, drains and isolation |
Plant-room layout | Equipment clearances, headers, supports, maintenance and removal access |
Coil connection details | Valve package, strainer, vent, drain, P/T ports and insulation |
Pipe-sizing schedule | Segment, flow, size, ID, velocity and friction |
Pressure-loss schedule | Critical circuits, fittings, equipment losses and totals |
Pump schedule | Duty, type, efficiency, NPSH, motor, VFD and redundancy |
Valve schedule | Type, size, pressure class, Cv/Kvs, flow, actuator and setting |
Control schematic/points list | Sensors, setpoints, interlocks, alarms and staging |
Specification | Materials, insulation, supports, cleaning, testing, TAB and water treatment |
BOQ | Pipes, fittings, valves, insulation, supports, equipment and instrumentation |
Testing, Flushing, Balancing and Commissioning
Commissioning requirements must be planned during design so the installed system includes the necessary access, drains, vents, measuring points and isolation.
Pre-commissioning
Confirm installation against approved drawings and valve schedules.
Pressure-test in accordance with the governing code and specification without exceeding the weakest component rating.
Flush and chemically clean the system using an approved method.
Clean strainers and verify that temporary strainers are removed when required.
Fill with treated water or the specified glycol mixture.
Vent all high points and verify the expansion-vessel precharge and fill pressure.
Pump and plant checks
Confirm pump rotation, alignment, lubrication and vibration.
Record suction and discharge pressure, motor current, speed and power.
Verify chiller evaporator flow and safeties.
Test duty/standby and multi-pump staging.
Prove isolation and check-valve operation.
Verify NPSH-related suction conditions and minimum-flow protection.
TAB and controls
Verify design flow through every terminal and chiller.
Set balancing valves or confirm PICV settings and operating differential pressure.
Establish the remote DP setpoint from measured system requirements.
Prove VFD control at full and part load.
Test every chiller staging and de-staging transition.
Prove controlled bypass operation at low load.
Test sensors, alarms, interlocks and failure modes.
Trend and handover
Trend at minimum:
CHWS and CHWR temperatures
System delta-T
Total and individual chiller flows
Pump speed, DP and power
Coil-valve positions
Chiller status and load
Bypass-valve position
Record final valve settings, measured flows, control setpoints, test results and as-built changes. CIBSE emphasizes planning commissioning provisions during design; see its water-system commissioning guidance.
Mistake | Likely consequence | Correct approach |
Adding building height to closed-loop pump head | Oversized pump and wasted energy | Use the critical-circuit dynamic losses; use height for fill pressure and ratings |
Applying diversity twice | Undersized plant flow | Keep room, coil and coincident plant-load boundaries clear |
Using water properties for glycol | Incorrect flow and head | Use actual density, specific heat and viscosity |
Selecting pipes only by velocity | Excessive friction or poor economics | Check both velocity and Pa/m plus life-cycle energy |
Assuming the farthest coil is critical | Insufficient pump head | Calculate all likely high-resistance circuits |
Treating reverse return as self-balancing | Unequal terminal flows | Still select and commission balancing/control valves |
Using uncontrolled bypasses | Low delta-T and excess flow | Use controlled minimum-flow protection only when required |
Ignoring valve authority | Unstable control and hunting | Select Cv/Kvs or PICV from actual differential pressure |
Oversizing pump safety margin | Noise, overflow and energy waste | Add only a documented, justified allowance after detailed calculation |
Locating DP sensor by distance alone | Starved critical branch or excess pressure | Place near the hydraulically critical area |
Incomplete vapour barrier | Condensation and corrosion under insulation | Seal all joints, supports, valves and penetrations |
No commissioning provisions | System cannot be measured or balanced | Design drains, vents, P/T ports, meters and access from the start |
Low delta-T is often a system problem rather than a chiller problem. Read Nexora’s guide to understanding low delta-T syndrome in chilled-water systems.
Frequently Asked Questions
1. How do you calculate chilled-water flow from cooling load?
For water, use Flow (L/s) = Load (kW) ÷ [4.186 × ΔT (K)]. Use actual fluid properties for glycol mixtures.
2. What CHWS and CHWR temperatures should be used?
There is no universal pair. Select temperatures with the chiller, coils, humidity-control requirement, energy strategy and any district-cooling contract.
3. What velocity should be used for chilled-water pipes?
Use a project-specific range and check friction, noise, pipe material, erosion, air management and pumping energy. Preliminary comfort-system screening often falls around 0.6–2.4 m/s depending on pipe size and service, but that is not a code rule.
4. What friction rate should be used?
Many designs screen within roughly 80–350 Pa/m, with some recognized guidance extending higher for selected conditions. The correct limit depends on life-cycle energy, acoustics, pipe material and project requirements.
5. Does building height increase pump head?
Not in a sealed closed loop at steady state; supply and return static heads cancel. Height does increase fill pressure, expansion-vessel precharge and component pressure rating.
6. How is the critical circuit identified?
Sum pipe, fitting, coil, valve, strainer and plant losses along each likely path. The path with the greatest total loss at design flow is the critical circuit.
7. Is reverse-return piping automatically balanced?
No. It makes circuit pipe lengths more nearly equal, but different coils, valves and branches still produce different pressure losses.
8. Where should the expansion vessel connect?
Normally at the point of no pressure change near the pump suction, subject to the approved system arrangement and manufacturer guidance.
9. Where should the differential-pressure sensor be located?
Near the hydraulically critical terminal branch or representative critical area—not simply at the physically farthest coil.
10. How much pump-head safety margin should be added?
There is no universal percentage. Complete the detailed loss schedule first, then add only a documented allowance appropriate to the design stage and remaining uncertainty.
11. When is variable-primary flow appropriate?
When the chillers permit the required flow range and rate of change, flow can be measured and controlled, staging is coordinated, and minimum-flow protection is proven.
12. Why does a chilled-water system develop low delta-T?
Common causes include excess flow, leaking or oversized control valves, three-way bypass flow, dirty coils, poor air balance, incorrect setpoints, uncontrolled bypasses and coils selected for a different temperature regime.
Final Design Checklist
Before issuing the design, confirm:
Cooling and coil loads are correct and diversity is applied once.
CHWS, CHWR and delta-T match selected chillers and coils.
Water or glycol properties are correct.
Every pipe segment has an assigned flow and size.
Velocity and friction meet approved project criteria.
Actual internal diameters and fitting quantities are used.
Equipment pressure drops come from current selections.
The critical circuit is calculated—not guessed.
Closed-loop static height is excluded from operating pump head.
Pump curve, efficiency, motor, NPSH, shut-off head and VFD range are verified.
Chiller minimum flow and staging transitions are protected.
Valves have adequate authority, close-off pressure and commissioning access.
Expansion, pressurization, air removal, dirt removal, vents and drains are complete.
High-rise pressure ratings and zones are checked.
Insulation and vapour barrier are continuous.
Supports, anchors, guides and movement are coordinated.
Controls, TAB, trending and handover requirements are written into the design.
Recommended Nexora Engineering Resources
Resource | Best used for | Access |
Pump Head Calculator Design Suite v1.0 | Pipe, fitting and equipment pressure losses; pump duty; pressure; motor-power estimate; NPSH check; printable report | |
Cooling Load Design Handbook | Cooling-load boundaries, psychrometrics, airflow, coil duty and equipment-selection strategy | |
HVAC Beginner’s Guide | Structured HVAC fundamentals, chilled-water systems, piping, valves, controls and commissioning |
Conclusion
A successful chilled-water piping design is a chain of verified decisions:
Cooling load → design delta-T → water flow → system arrangement → pipe sizing → pressure losses → critical circuit → pump selection → valves and controls → pressurization → commissioning.
The most common failures happen when one link is skipped—for example, selecting a pump before calculating the critical circuit, specifying a high delta-T without compatible coils, or designing variable flow without protecting the chiller’s minimum flow.
Follow the complete workflow, document every assumption and use current manufacturer data. The result will be easier to commission, more stable at part load and less expensive to operate.
Technical References
Engineering disclaimer: The equations, preliminary ranges and example values in this article are educational. Always verify the final design against applicable local codes, authority requirements, project specifications, approved equipment submittals, water-treatment requirements and manufacturer data. A qualified engineer must review and approve the project design.
Comments