Inflection Point Engineering Knowledge Base

Compressor Anti-Surge Control Strategy

Practical decision guide for when and how to apply anti-surge control on centrifugal and axial compressors. Covers single-service vs multi-service loops, invariant coordinate systems (ITCC), and the setpoint-ramp tricks that separate a compressor that never trips from one that trips weekly.

Why anti-surge control exists

A centrifugal compressor surges when the developed head can no longer overcome the downstream pressure. Flow momentarily reverses through the wheel, thermal growth differentials let rotor and diaphragm rub, and within seconds you have bent shafts, destroyed seals, and six-figure damage. The job of the anti-surge controller is to recycle or blow down enough gas to keep the operating point at least one safety margin to the right of the surge line, at all times, including startup and trip transients.

Three strategies ranked

StrategyWhen to useProsCons
Fixed minimum-flow recycleSmall single-stage air or nitrogen booster; constant suction conditionsSimple; no model requiredWastes recycle power at higher speeds / off-design operation
Flow / head control with parabolic surge lineSingle-service centrifugal; well-characterized family of curvesTracks curve; setpoint can be field-tunedNeeds good curve data; doesn't handle MW or T changes gracefully
Invariant-coordinate (ITCC / CCC Series 3+)Multi-service, multi-stage, variable MW, LNG trains, FCC wet-gasSurge line collapses to a single curve independent of P1, T1, MW, k; ramp-open on disturbanceRequires vendor software, commissioning time, special PT/FT transmitters on suction

Decision matrix: which transmitters do you need?

Tuning the setpoint

The static surge margin (SM) from a single operating point is not the setpoint — it's the target the controller must achieve during steady state. The actual control-line offset must add a transient margin for disturbance rejection.

LayerTypical % flow margin from surgeAction
Steady-state design10-15%Design basis; sets recycle-valve Cv
Control line (CL)6-10%Proportional-only opening of recycle valve
Trip line (TL)3-5%Ramp open to full stroke; CCC "T1 logic" or equivalent
Surge detection0%Open recycle 100% for minimum hold time; log event

Common pitfalls that cause surges despite an anti-surge controller

  1. Recycle valve too slow. Specify less than 2 s stroke time full travel for globe anti-surge valves (API 617 9th Ed. §6.10 recommends ≤1 s for large horsepower trains). Air-to-open with fast pneumatic positioners is standard.
  2. Recycle valve too small. Must pass 100% of maximum relieving flow with cold gas and at maximum DP. A common sizing miss is to use the steady-state suction flow rather than the startup or trip flow, which is 1.5-2x higher.
  3. Heat-recirculation on long recycles. If recycle returns to suction without aftercooling, repeated recycle events can walk the suction temperature up, shrinking surge margin. Add a dedicated recycle cooler or tie back to upstream KO drum.
  4. Instrumentation sampling too slow. DCS scan rates > 500 ms miss the surge event entirely. Surge control should live in a dedicated controller (CCC, Rockwell PlantPAx AS function block, Siemens SIMATIC PCS 7) with ≤ 50 ms scan.
  5. Loop interaction with pressure control. If discharge PIC and anti-surge both manipulate the recycle valve, they fight. Solution: override selector so PIC governs above the control line, anti-surge governs below.

Startup and shutdown logic

Recycle valve should open first on startup (full-open at zero speed), close as flow rises above the control line. On trip, recycle ramps open at ≥ 100%/s and stays open for a programmed hold time (typ. 60-120 s) to let the rotor coast down safely through surge.

When not to bother

Field diagnostic: is it surging?

If discharge pressure pulses at 2-10 Hz, shaft vibration spikes coincident with pressure dips, and you hear a distinct woof-woof at the suction strainer -- you're surging. Open the recycle manually immediately and reduce load.

References