Rotating equipment, not just compressors
The constraint changes. The architecture doesn't. A compressor's limit is aerodynamic stall; a pump's is hydraulic recirculation below minimum continuous stable flow. Both are protected by a recycle valve that spills flow back to suction. Both cost real energy every time that valve opens. And in both cases, the decisions that determine how often it opens are made a layer above the protective controller — where nobody is optimizing.
The constraint
Three limits define the operating envelope of a pump station. We treat all three conservatively.
Minimum continuous stable flow (MCSF)
Below it, suction and discharge recirculation drive radial thrust, shaft deflection, and seal and bearing damage. Protected by a minimum-flow recycle or ARC valve — the structural twin of the anti-surge recycle valve, and the same economic problem: pumped twice, delivered once.
NPSH margin
The cavitation constraint, and typically as sensor-poor as the equivalent constraint on the compression side. We treat it the same way: conservative envelope, widened when measurement quality degrades. We do not claim to shave NPSH margin.
BEP deviation
Running far from best efficiency point costs energy and reliability at the same time. Clean, measurable, and continuously drifting as the system curve changes.
The levers
The same supervisory decisions as compression, in liquid form.
| Lever | Notes |
|---|---|
| Parallel pump staging | The direct analogue of unit sequencing, and often higher value — pump stations frequently run more units, and affinity laws make the power curve steeply non-linear in speed |
| VFD speed vs. control-valve throttling | Throttling against a partly closed valve is destroyed energy. Where VFDs exist, the optimal split between speed and valve position drifts with system curve, viscosity, and temperature — and is almost never maintained |
| Multi-station hydraulic coordination | Choosing discharge pressures across successive mainline stations to minimize total power subject to MAOP and slack-line constraints |
| DRA dosing vs. pump horsepower | On drag-reducing-agent lines, chemical cost against electrical power is a real economic trade usually set by rule of thumb |
| Price and demand-charge response | Electric drivers allow optimizing when volume moves against pool price and demand charges, within linefill and delivery commitments — a lever gas-driven compression doesn't have |
| System-curve drift detection | Fouling, wear-ring clearance opening, and impeller wear appear as slow drift, which is what the adaptive approach is built to catch |
Unchanged
Unchanged: minimum-flow recycle and ARC control, low-suction-pressure trip, high-discharge and MAOP protection, surge-relief (waterhammer) systems, API 670 machinery protection, and API 682 seal system alarms. Same relationship as on the compression side — OmniPath sits above the protection layer and writes setpoints into the regulatory layer.
Works with your existing stack. OmniPath reads the midstream measurement layer — Emerson ROC and FloBoss, ABB Totalflow — and the SCADA historian, and writes setpoints through OPC UA to the regulatory layer. No controller firmware changes, no I/O changes, no safety system changes.
Where it applies
Liquids mainline pump stations, NGL and Y-grade lines, water injection and disposal, produced water handling, refinery charge and reflux service, and utility and cooling water pumping.