Variable Primary Flow Chiller Logic: Staging, Min-Flow Bypass, and Why It Beats Constant-Speed Plants

A primary-only-variable (P-O-V) chiller plant has lower pump kW, simpler controls, and lower capex than primary-secondary. But it requires intelligent chiller staging — when to add a chiller, when to drop one, when to engage the minimum-flow bypass. This logic is the difference between a P-O-V plant that saves 30% pump energy and one that low-flow-trips on the chiller daily.

This guide covers the staging logic, minimum-flow bypass design, and the BMS programming that makes it work.

Why P-O-V is the new standard

For new Indian commercial chiller plants > 1,500 TR, P-O-V is increasingly the default architecture. The pump-energy benefit is substantial:

Architecture Pump kW (1,500 TR plant) Annual energy Annual cost
Primary-secondary (CSP + VSP) ~150 kW peak; 90 kW avg 770 MWh ₹77 lakh
Primary-only-variable (VSP only) ~85 kW peak; 50 kW avg 410 MWh ₹41 lakh

Annual saving: ~₹36 lakh for a typical 1,500 TR plant.

But this saving only materializes if the staging logic works. Let’s cover what that means.

Chiller minimum flow

Every centrifugal/screw chiller has a minimum flow specification. Below this:

  • Refrigerant evaporator behaves erratically (sub-cooling errors)
  • Compressor surges (centrifugal)
  • Bearing wear (vertical orientation issues)
  • Eventually trips on low-flow alarm

Modern chillers handle 50-60% of design flow without issue. Older chillers required 80%+.

For typical 500 TR centrifugal chiller at 5 K range:

  • Design flow: 213 m³/h
  • Manufacturer minimum: ~108 m³/h (50%)
  • BMS conservative trip point: 130 m³/h (61%)

Minimum flow bypass valve

When building demand drops below chiller minimum, P-O-V plants engage a bypass valve. Excess primary flow circulates back to chiller return without passing through loads.

The bypass valve sized for: chiller minimum flow minus building demand at the lowest building load condition.

For 500 TR chiller at 50% min flow: bypass capacity ≈ 50% of design flow = ~107 m³/h.

The bypass valve modulates inversely to building load: at 100% load, bypass closed (all flow to building); at lower loads, bypass opens to maintain chiller minimum.

Chiller staging logic

The BMS decides when to add or drop chillers based on:

Add chiller (stage up):

  • Chilled water leaving temperature > 0.5 °C above setpoint for 5 minutes
  • OR currently-running chiller has been at 100% capacity for 10 minutes
  • OR primary flow is approaching maximum capacity (one chiller can handle)

Drop chiller (stage down):

  • Chilled water leaving temperature meets setpoint within ±0.5 °C
  • AND running chillers operating at < 50% combined capacity for 30 minutes

Stage-up logic for typical 1,500 TR plant (3 × 500 TR):

  • Cold morning startup: 1 chiller engaged
  • Mid-morning: load rises, ΔT widens → chiller 2 engages
  • Afternoon peak: load full, both chillers at design → chiller 3 engages
  • Evening: load drops → chillers 3, 2 drop in reverse order

Why staging logic fails

Common failures:

1. Hunt-and-peck staging

Chiller adds → demand drops → staging removes → demand returns → adds again. Cycles every 10-15 min.

Cause: Stage-down threshold too aggressive (e.g., < 60% combined load triggers drop).

Fix: Stage-down at < 50% AND for ≥ 30 min sustained.

2. Min-flow bypass never closes

Designer-set min-flow setpoint matches chiller capacity exactly. As building demand approaches chiller minimum, bypass partially open → flow inefficient.

Cause: No deadband around stage-up threshold.

Fix: Add 10% deadband: stage-up at “design – 10%” so bypass closes during stage-up.

3. Stage-up triggers when chilled water reset is active

Chilled water reset (raising target temp during off-peak) makes leaving temp acceptable at low flow. If staging logic uses fixed setpoint without acknowledging reset, false stage-ups.

Fix: Coordinate stage logic with chilled water reset schedule.

4. Chiller staging doesn’t match plant minimum capacity

4 × 500 TR plant in P-O-V; building demand drops to 200 TR. Single chiller still at 40% (below min flow). BMS doesn’t know to engage bypass + maintain capacity.

Fix: When demand falls below minimum chiller capacity, BMS engages bypass + holds one chiller on.

Worked example: 1,500 TR plant in Mumbai

Plant: 3 × 500 TR centrifugal P-O-V; 2-pump set on each chiller (duty + standby); bypass valve 200 m³/h capacity.

Hourly profile (Tuesday in summer):

Hour Building load (TR) Chillers running Bypass status Pump kW
06:00 200 1 (low) 50% open 35
08:00 700 2 Closed 105
12:00 1,400 3 Closed 165
18:00 800 2 Closed 110
22:00 250 1 50% open 40

Daily pump energy: ~1,700 kWh

Vs primary-secondary equivalent: ~2,400 kWh

Saving: ~700 kWh/day = ₹0.7 lakh/month = ₹8 lakh/year

Five common P-O-V design mistakes

1. No min-flow bypass valve. Building demand drops below chiller min → low-flow trip → cooling loss.

2. VFD on pumps but no VFD on chiller compressor. Chiller still runs full-speed on partial demand → wasted compressor energy.

3. No staging deadband. Hunt-and-peck behavior; chiller wear.

4. Bypass valve undersized. Cannot deliver chiller minimum at lowest building load.

5. No chilled water reset. Misses opportunity for free-cooling at part-load.

Quick checklist

  • [ ] Chiller minimum flow specified per manufacturer
  • [ ] Bypass valve sized for: max chiller flow – min building demand
  • [ ] BMS staging logic coordinated with chilled water reset
  • [ ] Stage-up + stage-down deadband (10% + 30-min sustained)
  • [ ] Differential pressure controller for VFD speed
  • [ ] Annual commissioning verification of staging behavior

References: ASHRAE Handbook HVAC Sys & Eqp 2024 Ch 13 (Hydronic Heating + Cooling); ASHRAE 90.1-2022 §6.5.4 (Pump Performance); ISHRAE Handbook 2024 Vol 4.

[Disclosure block, Legal notice — auto-included by article template]

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top