Pull the gland and you see a polished band running the full ID of the stuffing box bore, wider than the packing stack itself, with no matching wear on the plunger. That is not a packing material problem and it is not plunger runout. The rings are walking axially in the bore because the cavity behind them loses positive pressure on every suction stroke, and Kevlar packing sliding against 316SS eats the bore. Fix the suction system and the wear stops; bore-and-chrome without fixing suction buys you a longer bandage.
Read the wear pattern first
The wear geometry tells you which failure mode you have. Start with the bore, not the packing.
| What you see | What it means |
|---|---|
| Radial wear around the full bore ID, band wider than the ring stack, plunger clean | Packing rings shifting axially in the box. Loss of restraining pressure on the suction stroke. |
| Wear favoring one side of the bore, heavier at one clock position | Module-to-rod misalignment or bent piston/intermediate rod. |
| Heavy scoring on the plunger OD, bore relatively clean | Abrasive ingress or packing overtightened / run dry. |
| Stuffing box running hot, wear plus discoloration | Misalignment or excessive gland load, not suction. |
| Pitting on valve seats, pump sounding like a rock crusher | Inadequate NPSHa. Fluid column separating and reattaching in the cylinder. |
Full-circumference wear with a clean plunger rules out misalignment as the primary driver. A bent rod or a mislocated suction/discharge module drags the packing to one side and shows up as heat and a one-sided pattern. Nine pumps, same build, same packing, and only the two lines with bad suction piping failing at two weeks while the third runs three years, is a piping answer, not a component answer.
The mechanism: why suction pressure holds the packing still
Spring-loaded packing in a threaded stuffing box is held in place by two forces. The spring provides the preload. Cylinder pressure acting on the inboard face of the stack provides the rest, and it is the larger of the two once the pump is running. On the discharge stroke the stack is pinned hard against the gland. On the suction stroke that pressure drops toward suction pressure.
With 25 PSI of clean accumulator pressure at the manifold and good piping, the cavity stays positive through the whole suction stroke and the stack never unloads. The acceleration head required to start the fluid column moving on each suction stroke is proportional to the length of the column and the square of pump speed, and every fitting adds effective length. When acceleration head plus friction loss exceeds your available 25 PSI, cylinder pressure goes below suction manifold pressure and can go sub-atmospheric. Cavities open, then collapse when the discharge stroke starts.
That gives you two things at once. Valve seat pitting from bubble collapse, which you already have. And a per-stroke pressure reversal across the packing stack: the stack unloads, the spring is the only thing left holding it, and then the collapse event slams it back. At pump speed that is thousands of axial reversals per hour. Kevlar is dimensionally stable and hard on metal, so the box loses material instead of the packing. A band a centimeter wider than the ring stack is exactly what an oscillating stack produces.
Cavitation does not "migrate back" to the stuffing box as a fluid phenomenon. The suction deficiency that causes cavitation at the valves is the same deficiency that unloads the packing. Both are symptoms of one cause.
Run these checks before you touch the box
- Measure suction pressure at the pump manifold with a fast-response transducer, not a bourdon gauge. A dial gauge averages away the trough you are looking for. Log a full revolution and find the minimum.
- Compare that trace against the good line. Same instrument, same location, same rate. If the bad lines dip toward zero or negative while the good line stays positive, you are done diagnosing.
- Confirm the acoustic signature. Reciprocating pumps starved on suction jack and knock like a rock crusher. Centrifugal-style cavitation whine is not what you are listening for.
- Check the plunger for runout at zero, mid, and full stroke. Clean, and you have already eliminated a bent rod.
- Pull the packing and gauge the bore with an inside caliper at four points around the periphery, from the rod outward. Equal readings confirm the suction/discharge module is located correctly on its pins. Unequal readings are an OEM module replacement, and no amount of suction work will fix it.
- Check crosshead slide wear-down at the same three stroke positions, plus main, big-end, and crosshead pin clearances against the OEM settings. Multiple worn clearances stack into stuffing box trouble even when none of them alone would.
- Inspect the valve seats. Pitting on the bad lines and clean seats on the good line corroborates the suction diagnosis independently.
Fix the suction, in this order
Cheapest test first, permanent fix second.
- Raise suction pressure. If the system is sealed, increase accumulator pressure and watch whether the knocking stops and the minimum manifold pressure stays positive. This is the one-hour test that proves the theory before you spend money on pipe.
- Install an inlet stabilizer close-coupled to the manifold. A suction stabilizer supplies the acceleration head locally so the pump is not trying to accelerate the entire piping column each stroke. Mount it as close to the suction manifold as the flange pattern allows. Distance kills its effectiveness.
- Then repair the boxes. Bore and chrome once the suction is corrected. Chroming a box that still sees per-stroke unloading only slows the wear rate.
- Reconsider the packing last. Kevlar is aggressive on 316SS. Once the axial motion is gone the material choice matters far less, but a softer jacket over a chromed bore is the conservative pairing.
Verify the fix held
- Minimum suction manifold pressure stays positive through the entire suction stroke on the transducer trace, on all three pumps of the repaired line.
- Knocking is gone at full rate. If the pump only quiets down at reduced rate, the acceleration head is still marginal.
- Flow and discharge pressure hold at 90 GPM and 2900 PSI without the rate wander that pressure-starved cylinders produce.
- Pull one box at 30 days. The bore band should show no measurable growth. If it has grown, the suction is still deficient.
- Leak rate. You were at roughly a gallon per hour after two weeks. At 30 and 90 days a corrected line should be at drip-only.
Pitfalls that recur on this class of pump
- Chasing packing material. Every packing change buys a few weeks and hides the fault for another cycle. If one identical line runs three years, the packing is not the variable.
- Trusting a dial gauge on the suction. It will read 25 PSI while the cylinder is going negative twice a second.
- Mounting the stabilizer at the end of a run. Anything more than a short spool from the manifold and you are still accelerating most of the column.
- Overtightening the gland to stop the leak. On a spring-loaded stack that adds heat and accelerates bore wear without restoring the missing hydraulic restraint.
- Ignoring accumulated clearances. Worn crosshead slides and big-end clearance do not cause full-circumference bore wear, but they widen the band once suction has started the rings moving.
- Fixing one line and assuming the other follows. The two bad lines may not have identical piping faults. Instrument each one.
FAQ
Can cavitation on the suction side actually cause stuffing box wear?
Not directly, but the same suction deficiency does both. Insufficient NPSHa lets cylinder pressure drop below suction pressure each stroke, which unloads the packing stack and lets it shift axially while it also collapses vapor cavities at the valve seats.
Does full-circumference bore wear mean the pump is misaligned?
No. Misalignment produces one-sided wear plus elevated stuffing box temperature. Wear around the entire bore ID with a clean plunger points to axial ring movement from a suction-side pressure problem.
Can I just raise suction pressure instead of repiping?
Try it first as a diagnostic. If raising accumulator pressure stops the knocking and holds the manifold minimum positive at full rate, you have confirmed the cause, but the acceleration head from 90-degree elbows and hard tees near the manifold usually still needs piping work for a permanent fix.
Does boring and chroming the stuffing box solve the problem?
Only the symptom. Chrome slows the wear rate against Kevlar packing, but the rings keep oscillating until the suction system is corrected. Do the chrome work after the suction fix, not before.
Escalate to National Oilwell if the inside-caliper readings around the stuffing box periphery come back unequal, since relocating a pinned suction/discharge module is OEM work. Also involve them if bearing and crosshead clearances fall outside the OEM settings, because reconditioning the power end is not a field repair on a 200 HP frame.