O&M panorama: reading the paper machine as a conveyor of water and heat
Understand how the whole machine runs before discussing how to repair it.
1. One production line, five dewatering stages
What a paper machine does in one sentence: turn 1% consistency stock into 95% dryness paper. But that sentence runs through five dewatering stages, each with completely different equipment form, energy structure, and failure mode — so the first lesson of O&M is not to tear down equipment, but to read the line as a ‘dryness map’.
Take graphic paper: stock leaves the headbox at only 0.3%–1.2% consistency (depending on grade and approach system), rises to 18%–22% at the couch exit, 42%–48% at the press exit, 92%–95% at the dryer-section exit, and closes at 6%–9% finished-sheet moisture around sizing/calendering. Converted to ‘water’ magnitudes: per tonne of finished paper, the wire section removes over 97 tonnes of water, the press squeezes out about 1 tonne, and the dryer section evaporates about 1.1–1.3 tonnes. This ratio determines a series of things: wire-section problems show as ‘retention and blockage’, press problems as ‘dryness and crushing’, dryer problems as ‘steam consumption and sheet defects’. This ladder has another use: it is the common language of on-site communication. ‘The sheet broke at the third dryer group’ and ‘the sheet broke at 88% dryness’ describe the same event, but the latter immediately narrows the scope to causes like ‘high shrinkage stress, over-dry edges’. So O&M staff should form the habit: when reporting a fault, include the dryness position, not just the equipment position.
2. The capacity account: one formula to rule them all
A paper machine’s instantaneous capacity has only three variables: speed, width, basis weight. Any capacity discussion — retrofit, speed-up, grade change — starts from this formula:
Example: 1,200 m/min, 8.2 m width, 90 g/m² — Q = 1200 × 8.2 × 90 × 60 / 1000 ≈ 53.1 t/h. With 8,000 effective annual hours and 88% overall efficiency, annual output ≈ 374,000 t. Reverse use: a customer wants 300,000 t/year at 90 g/m² — what speed is needed? Divide 300,000 by 8,000 h and 0.88 to get 42.6 t/h, then back-calculate v ≈ 963 m/min — this is the first-layer arithmetic of ‘can we take this order’.
Note the leverage of basis weight: at the same 1,200 m/min, dropping basis weight from 90 to 60 cuts output by a third, yet the per-tonne drying load (water to evaporate per tonne) is almost unchanged — so low-basis-weight high-speed and high-basis-weight low-speed impose completely different pressure on the dryer section. From the O&M view, changing basis weight is not tweaking one recipe parameter but reallocating the dewatering and evaporating load across the line.
Multiplying the three rates is what gives meaning; looking at any one alone invites ‘cooking the books’: reporting only availability lets you count run-below-speed as running; reporting only quality lets you count speed-reduced good-paper selection as quality. Industry mature levels for graphic and packaging machines are roughly availability 92%–96%, performance 95%–98%, quality 96%–99%, with OEE at 84%–90%. Below 80% signals systematic loss; above 92% often means the statistical caliber widened rather than the equipment improved.
| Loss layer | Typical source | O&M lever |
|---|---|---|
| Planned downtime | Fabric/felt change, planned maintenance, cleaning, grade change | Window merging, job-package prefab, parallel work |
| Unplanned downtime | Web break, equipment failure, instrument misaction, stock interruption | RCFA, inspection, spare readiness, interlock setting |
| Speed-reduction loss | Insufficient drying, low vacuum, quality speed limit, machine stability | Bottleneck identification, steam & vacuum balance, cross-machine control |
| Quality loss | Start-up rejects, splices, defect downgrades, roll ends | Standardized start-up sequence, defect tracing, rewinder reject strategy |
Pit 1: using design speed for capacity. Design speed is the mechanical and drive upper limit; actual long-run speed depends on drying capacity and the quality window, often 10%–15% apart. Using design speed for investment calculation halves the payback period.
Pit 2: wrong width. Headbox width, wire width, and reel width differ by the trim amount (20–50 mm per side and up); using wire width to compute output systematically overestimates by 1%–2%, and 1% at the hundreds-of-thousands-of-tonnes scale is not small.
Pit 3: treating basis weight as constant. The 2σ of cross-machine basis-weight distribution is typically 1.5%–3%, but paper is sold by area, so the thicker-than-average cross-machine band is pure cost; use the measured cross-machine mean, not the set value.
3. The structure of downtime: classify first, then talk reduction
‘Too much downtime’ cannot be executed as an instruction; it must first be broken into statable categories. The common on-site split is by ‘controllability’ into four tiers: planned (roll/fabric change, maintenance, cleaning), process (grade change, wire wash, stock fluctuation), equipment (mechanical/electrical/instrument failure), external (power loss, steam supply, raw material, downstream blockage). Mixed together, the four classes can never yield a handle.
After splitting, rank. Experience says downtime follows Pareto: the top three causes usually account for 60%–75% of all downtime. But watch two ‘statistical traps’: one is short stops go unrecorded — speed drops and web breaks under 5 minutes often aren’t ticketed, yet cumulatively they can exceed a major overhaul; the other is soft cause classification — recording equipment failure as ‘operator error’ or process fluctuation as ‘equipment aging’ looks good on paper but leaves no handle for improvement. So the first discipline of downtime statistics is: one plant-wide classification table, reviewed by someone not carrying the metric.
4. The speed chain: there is always only one bottleneck
The paper machine’s sections string into a speed chain: stock approach (jet/wire ratio, typically 0.95–1.05) → wire → press → dryer → sizing & calendering → reel → rewind, with the pulp line upstream and the finishing section downstream. Whenever any link’s capacity falls below the set speed, the whole line runs at the weakest link — that is the bottleneck, and at any moment there is only one.
Finding the bottleneck needs no guessing; three pieces of evidence suffice: one, buffer level — whichever stock chest/broke chest/finished-roll buffer keeps accumulating, the bottleneck is upstream of it; two, speed-reduction records — a persistent ‘speed limit’ signal in the DCS (steam valve near full open, vacuum at the floor, drive current at rated); three, the direction of operators’ complaints — they always know which section ‘can’t keep up’, only nobody translates it into engineering language.
Four common bottlenecks and their traits: insufficient drying (steam valve >90% open yet target dryness not reached, high hood exhaust humidity); insufficient vacuum (couch vacuum drops, poor wire dewatering, low post-couch dryness); press-limited (line pressure already at max, felt compacted, post-nip dryness can’t rise); stock/approach-limited (chest level, cleaner differential pressure, screening capacity). The four have completely different remediation paths, so qualify first, then talk investment.
Many machines’ bottleneck is actually not on the paper machine but in rewinding, sheeting, or wrapping: the rewinder’s roll-change cadence can’t keep up with the reel, slitter blade change takes too long, rolls wait for the crane after coming off, or insufficient finished-goods storage causes roll pile-up. These show up as ‘the paper machine is forced to slow down’, but the cause is outside the machine, so the machine side frantically checks drying and vacuum for half a year with no result. The judgment is simple: watch the WIP pile-up in the rewind and finishing sections when the machine slows — pile-up is the answer. Such bottlenecks usually cost very little to fix (add a buffer location, change the roll-change flow, stagger blade changes), with far higher ROI than equipment retrofit, yet are most easily missed by ‘equipment thinking’.
5. Energy per tonne: the load account of the utility systems
The paper machine is a big energy consumer, and the energy structure determines the O&M focus of the utility systems. Typical magnitudes for graphic paper: steam 1.3–1.6 t/t paper (1.1–1.3 t is net evaporation, the rest heating stock, size drying, and line losses), electricity 500–700 kWh/t (vacuum system and stock pumps over half), water 5–15 m³/t (good closed-loop machines can be lower). Newsprint and mechanical-pulp grades use more steam (mechanical pulp enters the dryer section wetter), while board, with higher basis weight and width, actually has lower electricity per tonne.
The O&M meaning of these three numbers: high steam use, first check post-nip dryness (each 1-percentage-point dryness gain cuts drying steam 4%–5%) and the condensate system (siphon blockage, failed steam trap lets the dryer pond water, sharply worsening heat transfer); high electricity, first check the vacuum system (vacuum-pump water temperature, line leaks, suction-roll seal-strip wear); high water use, check direct discharge of seal and cooling water.
| System | Energy share (order) | O&M focus |
|---|---|---|
| Vacuum system | Electricity 25%–35% | Pump water temp & vacuum, seal water, line leaks, panel blockage |
| Stock approach | Electricity 15%–20% | Pump efficiency, consistency stability, screen & cleaner DP |
| Dryer section | Steam >90% | Condensate discharge, dryer temperature difference, hood ventilation balance |
| Hood & heat recovery | Steam 5%–10% | Exhaust humidity control, heat-recovery fouling & air leak |
| Drive & auxiliaries | Electricity 10%–15% | Speed-chain matching, bearing & alignment, no-load loss |
Raising post-nip dryness from 45% to 48% reduces water entering the dryer section by about 12%, allowing 5%–8% more speed at the same steam capacity — the most common path to ‘zero-investment speed-up’. Conversely, blindly raising press line pressure for dryness brings lower felt life, sheet crushing, and more breaks; the account must be combined: each 1-percentage-point dryness gain cuts steam 4%–5%, but may cut felt life 10%, and whether it pays depends on the felt-price to steam-price ratio.
6. Position index: turning equipment inventory into a queryable structure
For the O&M system to land, there must be a ‘machine map’: a graphic-paper machine has around two hundred rolls, plus pumps, fans, valves, and cylinders, with moving equipment in the thousand-item range. Without an index, spare planning, inspection routes, and job packages rely only on veteran memory. So the final step of this lecture is to inventory equipment into a three-level index: position → component → part.
| Section | Roll count (order) | Critical parts & O&M focus |
|---|---|---|
| Wire section | 60–90 | Forming fabric, vacuum couch roll, suction-box panel, wire-guide bearing |
| Press section | 15–25 | Press roll covering, shoe-press sleeve & shoe device, felt |
| Dryer section | 60–100 | Dryer cylinder (incl. journal & rotary joint), felt-guide roll, dry fabric |
| Sizing & calendering | 10–20 | Size roll, calender roll (incl. variable crown), doctor |
| Reel & rewind | 10–15 | Drum, support roll, slitter, tension control |
Once the index is built, three things become simple immediately: inspection routes are arranged by position (not by equipment type, to avoid missed patrols); the spare-parts ledger is hung by position (same model used in five positions — then you know how many to stock); shutdown job packages are packed by position (drawings, spares, tools, acceptance criteria together). The specific practice of these three is spread across the maintenance-methodology, roll-specialty, and spare-parts lectures.
7. Ten numbers at the pre-shift meeting: twenty readings to watch the whole machine
The use of panorama knowledge finally lands on ‘which numbers to watch each day’. Experience says twenty numbers suffice to cover a graphic-paper machine’s running state, arranged by section; a pre-shift meeting run through them, and anomalies jump out by themselves — faster than flipping alarm logs, and more accurate than waiting for shift-handover verbal descriptions.
| Section | Numbers to watch | Normal range (order) & anomaly meaning |
|---|---|---|
| Approach | Stock consistency, jet/wire ratio, manifold DP | Consistency ± 0.05%; jet/wire 0.95–1.05; DP jitter = cleaner or screen blocked |
| Wire section | Couch vacuum, post-couch dryness, fabric tension, suction-box vacuum ladder | Couch 40–70 kPa; dryness 18–22%; vacuum ladder steps up, drop one level check one level |
| Press | Line pressure, post-nip dryness, felt permeability/wash DP, shoe-press oil temp & pressure | Dryness 42–48%; oil temp 45–55 ℃; oil pressure fluctuation = oil film or pump anomaly |
| Dryer | Each dryer group’s steam pressure & temperature difference, condensate temp, hood supply/exhaust wet-bulb | Inter-group ΔT <5 ℃; condensate temp near saturation; exhaust humidity 0.15–0.25 kg/kg |
| Sizing & calendering | Size amount, calender line pressure, roll-face temp, cross-machine thickness | Cross-machine thickness 2σ <1.5%; roll-face ΔT <3 ℃ |
| Reel & rewind | Reel tension, roll-change success rate, splice count, roll hardness | Splices <1/roll; roll-change failures tracked monthly, >2% check program & slitter |
| Utility | Steam per tonne, electricity per tonne, vacuum-pump water temp, white-water consistency | Steam 1.3–1.6 t/t; electricity 500–700 kWh/t; pump water temp <45 ℃ |
The value of these twenty numbers is in cross-verification: post-nip dryness dropped, steam use up, vacuum normal — the problem is most likely in the press (felt compaction or insufficient line pressure); post-nip dryness normal, steam use also up — problem in condensate or hood; both normal but speed can’t rise — look at drive and rewind cadence. A single number only alarms; linking them locates.
The same vacuum 55 kPa, yesterday 62, today 55, matters far more than the absolute ‘60 rated’. Every inspection sheet should carry two columns ‘last value’ and ‘change rate’, otherwise the reading is just ritual. Three empirical thresholds for trend judgment: a single-shift change over 5% warrants attention, three-day one-direction drift warrants checking, change accompanied by rising energy warrants a shutdown-window repair.
8. O&M differences across four paper grades (see differences first, then dive into volumes)
| Dimension | Graphic paper | Board / kraftliner | Tissue |
|---|---|---|---|
| Dryness difficulty | Formation & surface strength | Multi-layer bonding & moisture cross-machine | Creping & feel |
| Consumable density | Fabric, felt, doctor | Fabric, felt, size/coating system | Creping blade, coating, forming fabric |
| Availability caliber | By reeled finished product | By reeled finished product, incl. grade change | By wound finished product, frequent roll changes |
| Typical bottleneck | Drying capacity, vacuum | Multi-layer approach, press, dry end | Yankee drying & creping consistency |
| Spare-parts trait | Roll covering, vacuum parts | Many rolls + pulp-line grinding/screening parts | High-value roll sleeve, creping blade, coating chemistry |
The pulp line is a fourth category, but it belongs to the ‘fiber line’ not the ‘paper machine’: pulping, refining, screening, cleaning, washing, bleaching — completely different equipment form and failure mode (refiner plates, screen drums, pump-shell wear), covered in a dedicated lecture in Part C. Viewing the pulp line inside the paper-machine map has one meaning only: it is the upstream buffer of the paper machine, and chest-level and stock-quality fluctuations show up as ‘paper-machine-side web breaks and cross-machine variation’ while the maintenance report labels them ‘wet-end problems’.
The tissue machine deserves one specific note: it has only one Yankee dryer for drying, and evaporation capacity is strongly coupled to cylinder-surface coating and hood air velocity, so ‘speeding up’ on a tissue machine is often not a drying-capacity issue but a creping-consistency issue — if blade angle, coating state, and cylinder-surface temperature deviate slightly, the crepe won’t be uniform, and higher speed only yields more rejects.
Key takeaways of this lecture
- The paper machine is a dryness ladder: 0.3–1.2% → 18–22% → 42–48% → 92–95% → finished sheet 6–9%; the wire section removes over 97% of the water, press and dryer each handle about 1 tonne.
- Capacity Q = v·B·g·60/1000; 1,200 m/min × 8.2 m × 90 g/m² ≈ 53.1 t/h — speed-up, basis-weight change, and order-taking all start from this formula.
- OEE = availability × performance × quality; the three rates must be reported together; mature graphic-paper machines run OEE 84%–90%.
- Downtime is tallied in four classes — planned / process / equipment / external; the top three usually account for 60%–75%; the true cost of a web break must be counted by ‘total recovery time’, with dry-end re-ramp 3–5× the downtime duration.
- On the speed chain there is always only one bottleneck: first check buffer level, then speed-limit signals, finally ask operators; fixing the bottleneck makes it move — that is progress.
- Per-tonne orders of magnitude: steam 1.3–1.6 t, electricity 500–700 kWh, water 5–15 m³; each 1-percentage-point gain in post-nip dryness cuts steam 4%–5%, but felt life may drop 10% — whether it pays depends on the steam-price to felt-price ratio.
- Equipment management uses a three-level index: position → component → part; inspection routes, spare ledger, and job packages all hang on this index.
- Twenty numbers, one trend discipline: pre-shift readings must carry last value and change rate; a single-shift change over 5% warrants attention, three-day one-direction drift warrants checking, change with rising energy warrants a window — readings unlinked only alarm, they cannot locate.