One refinery in the Caribbean switched to Plan 32 for their light-cycle oil pumps, expecting better seal protection. Instead, they lost hundreds of barrels per day of refined gas oil — clean flush fluid migrating through the throat bushing into lower-value coker charge. Plan 32 solved the contamination concern and replaced it with a massive waste problem.
That case captures the decision most engineers get wrong. The Plan 11 vs Plan 32 choice is not a simple clean-vs-dirty binary. It is a cost-reliability tradeoff: Plan 32 eliminates process fluid contact at the seal faces, but it introduces external supply dependency, flush fluid economics, and operational complexity that many specifications never account for. Before defaulting to Plan 32, you need to know exactly what you are trading away.
Where Plan 11 Works — and Where It Stops Working
Plan 11 handles somewhere between 50% and 75% of all mechanical seal installations worldwide. That number surprises engineers who treat it as a basic-service-only option, but it reflects how many pumps run on reasonably clean, moderate-temperature fluids where recirculating process liquid from pump discharge provides adequate cooling and lubrication.
The operating envelope is wider than most assume. A 1/8-inch orifice at 100 psi differential delivers 2-3 GPM of flush flow — enough for standard single seals in water, light hydrocarbons, and chemical services below roughly 200-250 F. Stepping up to a 3/16-inch orifice pushes flow to 5-7 GPM for higher heat loads.
Where Plan 11 genuinely fails is when the process fluid itself becomes the problem. Three conditions disqualify it:
- The fluid carries abrasive solids that accelerate seal face wear faster than the flush can clear them.
- The fluid polymerizes or crystallizes at seal face temperatures, coating and destroying the running surfaces.
- The fluid’s vapor pressure at seal chamber temperature approaches or exceeds flush pressure, starving the seal of liquid film.
Before any of those disqualifiers apply, modifications can extend Plan 11’s range. Series orifices or choke tubes handle high differential pressure where a single orifice would generate excessive flow. For moderate solids, adding a cyclone separator (Plan 31) removes particles when solid density exceeds roughly twice the liquid density. The minimum orifice bore stays at 1/8 inch — go smaller and plugging becomes the failure mode, not contamination.
I recommend exhausting these modifications before specifying Plan 32. Every Plan 11 modification adds a few hundred dollars in hardware. Every Plan 32 installation adds thousands in piping, a continuous flush supply requirement, and an entirely new category of failure modes.
What Plan 32 Actually Costs You
For genuinely abrasive slurries, polymerizing fluids, or services where the process liquid cannot lubricate seal faces, Plan 32 or a pressurized dual seal is the only option. The problem is everything else that comes with it — flush fluid waste, supply dependency, and operational overhead that most specifications underestimate.
Flush Fluid Waste
Most Plan 32 installations are overdesigned by a factor of 10 to 20. Standard practice sets flush pressure 15-25 psi above seal chamber pressure, which often drives 5-10 GPM through the system. Properly engineered for actual cooling and purge requirements, many applications need less than 0.5 GPM. That gap matters because the flush fluid migrates through the throat bushing into the pumpage — it is consumed, not recirculated.
When the flush source is plant water, the waste is tolerable. When it is a refined product, solvent, or treated fluid, the economics change fast. I’ve seen plants where the flush fluid was genuinely more valuable than the pumpage it was protecting. Cost pressure then drives operators to restrict flow rates, which is exactly where the next problem starts.
Supply-Dependency Failures
Plan 11 is self-contained — if the pump runs, the flush runs. Plan 32 depends on an external supply system: piping, control valves, pressure regulators, sometimes a dedicated pump. A supply interruption of even a few seconds can irreversibly damage seal faces.
I’ve seen this failure pattern dozens of times on hot standby pumps. Operators restrict or shut off Plan 32 flush to idle pumps to save fluid. On startup, the seal runs dry for the seconds it takes to establish flow. After pulling those seals, you find contamination and corrosion on the seal head — clear evidence of inadequate flush during the standby period. The seal was not defective. The flush plan created a single-point dependency that operations could not consistently maintain.
For high-temperature services, Plan 32 introduces another trap. Engineers specify it for cooling benefit, but uncontrolled supply temperature can cause thermal shock on startup. Plan 32 is often misapplied to high-temperature applications where Plan 21 (cooled recirculation) would provide temperature control without supply dependency.
Plan 11 Disqualification Checklist
Stop evaluating Plan 32 advantages and start with this question: does your application specifically disqualify Plan 11? If none of these triggers apply, Plan 11 — possibly with modifications — is the simpler, more reliable choice.
Fluid contamination trigger. Abrasive solids at concentrations high enough to accelerate face wear, or particles that cause visible scoring on seal faces within your maintenance interval. If solids density exceeds twice the liquid density, try Plan 31 (cyclone separator) first. Sharp or crystalline particles usually overwhelm cyclone separation and require Plan 32.
Polymerization trigger. If the process fluid forms films, deposits, or polymer buildup at seal face operating temperature, no amount of Plan 11 flow prevents the damage. Check operating temperature against the fluid’s polymerization onset — this is the clearest Plan 32 disqualifier.
Vapor margin trigger. When seal chamber pressure minus fluid vapor pressure at operating temperature drops too low for a stable liquid film, Plan 11 starves the seal faces. Flash vaporization causes dry running. Plan 32 with a higher-boiling flush fluid is the correct fix.
Flush economics trigger. Even when Plan 32 is technically correct, verify your plant can sustain the flush supply economically. Calculate actual flow requirement (cooling flow vs purge flow, whichever governs), multiply by operating hours, and price the consumed fluid. If annual flush cost exceeds seal replacement cost under Plan 11 with shorter intervals, the economics may favor more frequent seal changes.
Supply reliability trigger. If your plant cannot guarantee uninterrupted flush supply — including during startups, shutdowns, upsets, and standby periods — Plan 32 will eventually fail.
The Bottom Line
Plan 32 is not an upgrade from Plan 11. It is a different set of operating commitments. Choosing Plan 32 means accepting supply dependency, flush fluid cost, and operational complexity in exchange for eliminating process fluid contact at the seal faces. For abrasive slurries, polymerizing fluids, and services with inadequate vapor margin, that trade is necessary and correct. For everything else, Plan 11 with proper orifice sizing or a cyclone separator modification handles the job with fewer failure modes and lower total cost. The API 682 standard specifies this progression for a reason — exhaust the simpler plan before adding complexity.
