One is supposed to leak and one is not. Everything else — cost, maintenance, water use, energy, reliability — follows from that single fact.
Packing MUST leak. That drip is lubrication and cooling.
Tightening it dry is the fastest way to destroy a shaft sleeve.
Compression packing has been the traditional means of sealing centrifugal pumps for more than 100 years.
| Component | Function |
|---|---|
| Stuffing box | The cavity around the shaft where the packing sits |
| Packing rings | Braided, rope-like, lubricated material cut into rings and stacked around the shaft |
| Gland follower | The adjustable plate that compresses the stack |
| Lantern ring | A spacer ring mid-stack, aligned with the flush port, that distributes seal water through the packing set |
| Shaft sleeve | A replaceable sacrificial surface. The packing rubs on this, not on the shaft itself. |
Compressing the rings expands them radially against the sleeve and the bore, creating a controlled leak path parallel to the shaft axis. That path is the seal, and it's deliberately imperfect.
| Component | Function |
|---|---|
| Rotating face | Turns with the shaft |
| Stationary face | Held in the seal chamber or gland plate |
| Spring or bellows | Holds the faces in contact, accommodates shaft movement and face wear |
| Secondary seals | O-rings or gaskets sealing the static leak paths behind each face |
| The seal itself | A microscopically thin fluid film between two lapped faces. That film both seals and lubricates. |
Packing seals along the shaft, and leaks along that path on purpose.
A mechanical seal seals at right angles to the shaft, between two flat faces, and is designed to leak as little as physically practical.
Every difference in cost, maintenance, water use, energy and reliability follows from that geometry.
"The packing's job is to reduce leakage. Not prevent it entirely." — Crane Engineering
| Function of the leakage | Consequence if you eliminate it |
|---|---|
| Lubrication of packing against the sleeve | Friction rises sharply |
| Cooling — carries away frictional heat | Packing burns and hardens |
| Flushing the interface | Solids embed and score |
| Net result of running dry | The packing burns and wears grooves into the shaft and sleeves |
An operator sees a drip, tightens the gland to stop it, and walks away satisfied.
What actually happened: friction went up, temperature went up, the lubricant embedded in the packing cooked out, and the sleeve started to groove.
A grooved sleeve cannot be sealed by any amount of packing. The repair is a sleeve replacement, which means pulling the pump. Trade literature describes under- and overtightening of packing rings as a prevalent and growing misapplication of centrifugal pump maintenance.
| Source | Recommended leakage |
|---|---|
| AESSEAL | ~1 drop/min per inch (25 mm) of shaft OD |
| Crane Engineering | 10 to 15 drops per minute |
| Magpie | 15 to 20 drops per minute |
| Dynapro | 10 to 20 drops per minute per stuffing box |
| Mislier | 40 to 60 drops per minute |
These aren't all describing the same pump. Required leakage scales with shaft diameter, surface speed and stuffing box pressure — a 6-inch shaft at 1,750 rpm needs far more cooling flow than a 1-inch shaft.
That's why the scaling formulation is the more defensible one: one drop per minute per inch of shaft gives roughly 2 dpm on a 2-inch shaft and 6 dpm on a 6-inch. The flat numbers are shorthand for common mid-size pumps and shouldn't be applied to a very small or very large shaft.
The answer that governs: your packing manufacturer's specification, for that packing, in that service, on that shaft. Get it in writing and post it at the pump.
The one number everybody agrees on is that zero is wrong. Every source consulted — including seal manufacturers with a commercial interest in packing looking bad — states that packing requires visible leakage to survive.
| COMPRESSION PACKING | MECHANICAL SEAL | |
|---|---|---|
| Design leakage | Required and visible | Near zero — millilitres per hour |
| Upfront cost | Low | Significantly higher |
| Maintenance labor | Routine gland adjustment, periodic repack | Essentially none once installed |
| Adjustment | Frequent, with no set degree of tightening | None |
| Failure mode | Leaks more. Adjustable. Pump usually stays in service. | Can require immediate shutdown |
| Shaft sleeve | Wears — a consumable | Minimal wear |
| Abrasive / fibrous service | Tolerant | Vulnerable without a proper flush plan |
| Hazardous or expensive fluid | Poor — continuous loss | The right answer |
| Housekeeping | Ongoing leakage to clean up | Dry |
| Field repairable | Yes, by most maintenance staff | Usually a replacement |
| Installation skill | Moderate | Higher — or use a cartridge design |
| Bearing risk | Leakage and washdown reach bearings | Lower |
"When packing fails it simply starts to leak more and so can be readjusted and tightened to reduce the leak with the pump remaining in service. Whereas mechanical seal failure can necessitate the shutdown of the pump. This is why pump packing is the primary source of packing for fire pumps and other applications that can't risk catastrophic failure."
That's AESSEAL — a mechanical seal manufacturer — making the strongest available argument for packing on their own website. The point is legitimate: packing fails gradually and gracefully. Seals can fail suddenly and completely. On an unspared critical pump, that matters.
Chesterton — which sells compression packing and mechanical seals — states: "although some say that packing requires more energy than mechanical seals, our testing has found that they are essentially the same."
That's a claim against interest on the seal side of their own business, which makes it worth taking seriously. It directly contradicts most of the marketing in this space.
| Source | Claim |
|---|---|
| Chesterton (sells both) | Essentially the same, based on their own testing |
| Magpie | Mechanical seal friction power is only ~10–50% of packing |
| Dynapro | Packing friction can add several percent to pump power demand, worse when hot or dry |
| BBP Pump | Mechanical seals consume less energy than gland packing |
| Stream Pumps | Gland packing creates higher shaft drag and greater energy losses |
The method costs nothing: watch motor kW before and after a packing adjustment, or compare kW for similar pumps with seals versus packing.
That single measurement settles the question for your pump, in your service, with your packing — the only version of the question that matters. Then translate it: incremental kW × operating hours × $/kWh = annual energy penalty.
The disagreement may be less about packing versus seals and more about properly adjusted packing versus overtightened packing.
Chesterton's testing presumably used correctly adjusted packing. The multi-percent penalties others cite align with the fact that overtightening restricts leakage flow and increases friction.
That reframes it usefully: the penalty isn't inherent to packing. It's a consequence of the maintenance practice around packing — which is harder to control than a seal that needs no adjustment at all.
| Scenario | Loss |
|---|---|
| One stuffing box at 60 drops/min | ~3 litres per hour |
| A 10-pump station running continuously | ~262,000 litres per year |
Mislier. That source uses the 40–60 dpm figure; at 10–20 dpm the loss would be proportionally lower.
Where packing is fed by a clean flush rather than by the pumped fluid, that water is potable or treated water going to drain — and in slurry service, it's also dilution of the process stream you're trying to thicken.
Maintenance staff commonly hose leaking product away from base plates. That moisture, plus leakage from failed packing, wrecks bearings.
AESSEAL cites a Mobil Oil study finding that water contamination of just 0.002 percent reduces rated bearing life by almost 50 percent. And Chesterton notes that packing removal almost always means changing the bearings.
So the true cost of a repack isn't the packing. It's the packing, plus the sleeve if it's grooved, plus the bearings, plus the labor to pull the pump.
| Packing is better when… | A seal is better when… |
|---|---|
| The fluid is abrasive, fibrous or full of solids — slurries, grit, raw sludge | The fluid is hazardous, volatile, expensive or environmentally sensitive |
| Unplanned shutdown is unacceptable and graceful failure matters | Leakage is a regulatory or emissions problem |
| Budget won't support seals and maintenance staff is available | Water conservation matters, or flush water is expensive to treat |
| The stuffing box is accessible for routine adjustment | Product dilution or contamination by flush water is unacceptable |
| Minor leakage is genuinely harmless | Maintenance staffing is limited — seals need no routine attention |
| The pump is old, low-value, or scheduled for replacement | The pump is high value, high pressure or high temperature |
Low-pressure clean-water centrifugal pumps in the 5–15 bar range with accessible maintenance can run either option. In that band the decision comes down to your staffing, your water cost, and how much you value the graceful failure mode. There is no technically wrong answer.
| Service | Consideration |
|---|---|
| Raw wastewater and grit | Abrasive. Packing tolerates it; a seal needs a proper flush plan and clean flush water. |
| Sludge and slurry | Same — plus the flush water dilutes the sludge you're trying to thicken, which is a real argument for seals here |
| Secondary effluent, filter feed, plant water | Clean enough for seals. Good conversion candidates. |
| Chemical feed | Seals — leakage is a safety and cost issue |
| Dry-run risk | A mechanical seal running dry fails fast. Where a wet well can go empty, that risk needs a control response. |
A fresh packing set is supposed to leak heavily at first. Start the pump with the gland finger tight and let it leak. A generous initial leak is correct.
Tighten gradually — a flat or a sixth of a turn at a time — with time between adjustments for temperature to stabilize. Feel the gland. It should be cool to warm, never hot. Heat means you've gone too far, and the correct response is to back off, not push through.
Work down to target leakage over the first hour or several hours, not the first minute. A packing set cranked down on startup is a packing set that will be replaced early — along with the sleeve.
| Ongoing task | Note |
|---|---|
| Check leakage | Daily on critical pumps |
| Adjust in small increments | There's no set degree of tightening — which is exactly why mistakes are easy |
| Feel the gland temperature | The best free diagnostic you have |
| Verify flush water | Loss of flush in abrasive service kills packing fast |
| Log adjustments | Gland travel over time tells you when a repack is due |
| Plan the repack | Rather than adjusting a spent set until the sleeve is ruined |
| Category | Options | Note |
|---|---|---|
| Configuration | Component vs cartridge | Cartridge designs reduce installation variability substantially |
| Arrangement | Single vs double (dual) | Double seals add a barrier or buffer fluid — hazardous fluids, dry-run protection |
| Balance | Balanced vs unbalanced | Balanced seals reduce face loading, allowing higher pressures |
| Spring type | Pusher vs non-pusher / bellows | Bellows avoid the hang-up and fretting that affect pusher seals in dirty service |
| Face materials | Carbon, silicon carbide, tungsten carbide, ceramic | Harder pairs for abrasive service; carbon for lubricity |
The faces are separated by a film thinner than a human hair. Grit in that film destroys them.
That's why standardized seal flush plans exist — recirculation from discharge, external clean flush, quench, barrier fluid systems. A seal conversion in wastewater or slurry service is not just a seal purchase. It's a seal plus a flush arrangement, and the flush arrangement is frequently where conversions fail.
Specify the flush plan with the seal. Don't treat it as an accessory.
On cartridge designs specifically: what you're buying is precision faces, springs, secondary sealing elements, and a design that reduces installation variability. On a component seal, correct face loading depends on the installer measuring and positioning correctly. On a cartridge, that dimension is set at the factory and locked with setting clips. For a plant with rotating maintenance staff, that matters more than face materials.
Many older stuffing box housings need a conversion sleeve or cartridge adapter to accept a mechanical seal. Confirm bore dimensions and shaft diameter before specifying.
Budget for the possibility that the conversion includes machining, a new sleeve, a seal chamber conversion and flush piping — not just the seal.
A maintenance culture built around packing has a reflex: see a drip, tighten something.
On a mechanical seal there is nothing to tighten, and a drip means the seal has failed. Tightening the gland plate bolts on a cartridge seal accomplishes nothing and can damage it.
Conversions fail for human reasons as often as technical ones. Say this out loud when the new seal goes in.
Shaft runout is the hidden dealbreaker. Packing is forgiving of runout, bent shafts and worn bearings because the braided rings deform to follow. A mechanical seal isn't — the faces must stay parallel and in contact. A pump that ran acceptably on packing for years may reveal a shaft or bearing problem the moment a seal goes in. Check runout before you convert, not after the new seal fails.
| Packing failure | Cause |
|---|---|
| Excessive leakage | Normal wear — lubricant depletes, ring volume falls. Adjust, then plan a repack. |
| Burned or hardened packing | Overtightening or loss of flush. Ran dry. |
| Grooved shaft sleeve | Overtightening, abrasives, or running dry |
| Extrusion | Excessive pressure or wrong packing selection |
| Won't hold adjustment | The sleeve is worn. Stop adjusting and repack. |
| Seal failure | Cause |
|---|---|
| Sudden heavy leakage | Face damage, cracked face, or a blown secondary seal |
| Face wear or scoring | Abrasives in the film. Inadequate flush. |
| Heat checking | Dry running or inadequate cooling |
| Fretting on the shaft | Pusher seal secondary element hanging up |
| Early failure after install | Installation error — the argument for cartridge designs |
| Failure after conversion | Frequently shaft runout, bearing condition or flush plan — not the seal |
| Error | Consequence |
|---|---|
| Tightening packing until it stops leaking | Burns the packing and grooves the sleeve. The single most common mistake in this subject. |
| Applying a flat drops-per-minute figure to any shaft size | Required leakage scales with shaft diameter. Use the manufacturer spec. |
| Cranking down a new set on startup | Skipping break-in guarantees early failure |
| Not staggering ring joints | Creates a straight leak path through the stack |
| Mispositioning the lantern ring | The flush never reaches the packing |
| Leaving old rings in the box | Common cause of early failure |
| Packing over a grooved sleeve | It will never seal |
| Hosing down leakage around the pump | 0.002% water contamination cuts rated bearing life nearly in half |
| Not budgeting bearings into a repack | Packing removal almost always means changing the bearings |
| Converting without checking shaft runout | Packing forgave it. The seal will not. |
| Not retraining staff after conversion | On a seal, a drip means failure and there's nothing to tighten |
| Accepting energy claims without measuring | A manufacturer selling both says the difference is essentially nothing |
| Item | Value |
|---|---|
| Packing design leakage | Required — visible drip |
| Leakage — scaling method | ~1 drop/min per inch (25 mm) of shaft OD |
| Leakage — flat figures cited | 10–15, 15–20, 10–20, 40–60 drops/min by source |
| Mechanical seal leakage | Millilitres per hour when correctly applied |
| Water loss at 60 dpm | ~3 L/hr per stuffing box |
| 10-pump station at 60 dpm | ~262,000 L/year |
| Bearing life vs water contamination | 0.002% water reduces rated life by ~50% |
| Packing failure mode | Gradual — adjustable, pump stays in service |
| Seal failure mode | Can be sudden — may force shutdown |
| Why fire pumps use packing | Graceful failure; cannot risk catastrophic loss |
| Energy difference | Disputed — measure your own kW |
| Ring joint stagger | ~90° (four rings) or ~120° (three rings) |
| Break-in | Start finger tight, tighten gradually, keep the gland cool |
| Ambiguous middle ground | Low-pressure clean water, 5–15 bar, accessible maintenance |
| Conversion risks | Shaft runout, flush plan, bore dimensions |
Leakage rates range from ~1 drop/min per inch of shaft to a flat 40–60 dpm across sources. Presented as a spread. Your packing manufacturer's specification for your service governs.
The energy comparison is genuinely disputed — including by a manufacturer that sells both products. Measure motor kW at your own plant rather than relying on any published figure, including these.
Every source here is a seal manufacturer, packing manufacturer, pump distributor or trade publication. All have commercial interests. Claims against interest — AESSEAL on packing's advantages, Chesterton on energy — are given more weight for that reason.
Seal flush plan numbering and selection is a substantial subject requiring the applicable standard and the seal vendor's input. Not covered in detail here.
Installation guidance is general good practice. Follow the packing and pump manufacturer instructions, which govern. Seal and packing selection for a specific service should involve the pump OEM and the seal vendor.
Operating: packing has to leak, and the drip is doing work. Tightening it dry is how sleeves get ruined and how a cheap consumable turns into a pump teardown.
Buying: the graceful failure mode of packing has real value on unspared critical pumps — and a seal manufacturer will tell you so.
The Complete Exam Guide covers all 12 major exam topics with 200 conceptual questions and a detailed explanation behind every answer — including pumps, mechanical systems and maintenance.
Get the Complete Exam Guide — $17Instant PDF download · 200 questions · No math · All topics