Paper, Wood and Packaging Manufacturing Case Studies
Industrial5 Case StudiesUpdated Sep 2026
Paper, wood and packaging plants run on tight margins and long equipment lives, so one failed component or a small drift in board caliper can disrupt a full week of orders. The five cases below come from fictional sites and cover situations common in this sector: a failing press roll, a bottlenecked carton line, a safety review on a waste baler, a late press delivery, and an energy program at a board mill. Each follows the same frame of situation, challenge, analysis, actions, and results. Treat them as working notes, since the figures belong to each site.
Case Study 1
Repeated failure of a press roll bearing housing
Kestrel Board Works, a single-machine linerboard mill that runs its board machine on two shifts.
The situation
Kestrel Board Works runs one board machine that produces about 4,200 tons of linerboard each month across two shifts. Over the last several months, the bearing housing on the lower press roll had cracked three times, and each failure forced an unplanned stop of about eleven hours. The housing is a cast iron part bolted to the frame, and it carries the roll load that keeps the sheet pressed evenly as moisture leaves the web. Each crack started on the inboard side and was found only after operators heard a change in the roll noise.
The challenge
The maintenance crew could swap the housing quickly, and they had already done that twice, so the spare part was not the real concern. Production planning needed to know whether the cracks came from a design weakness, from loads above the original specification, or from missed maintenance routines. Without that answer, any fix would be a guess, and a fourth unplanned stop before the outage would cost the mill a large share of the month's margin.
Analysis
The engineering review started with the constraints. The housing was sized for a nip load based on a specific basis weight range, and the mill had moved toward heavier sheets for a customer order during the previous quarter. Operating data showed roll load climbing past the original design envelope on the heavier grades, while caliper tolerance stayed tight, so the press worked harder to hold the same moisture profile. Maintenance records added a second layer. Lubrication intervals on the housing bearings had been extended twice to fit the shutdown calendar, and two alignment checks were logged as deferred and never closed.
Teams grounded in Paper, Wood and Packaging Manufacturing usually separate a failure into the part, the load, and the routine that let the load go unchecked. In this case the crack pattern pointed to fatigue starting at a casting radius, which fits loading above the rated envelope. The corrective options differ in cost and in what they leave unresolved. A redesigned housing with a larger fillet radius eases the stress concentration but leaves the load as it is. Capping the nip load protects the housing but may widen moisture variation at the reel. Restoring lubrication and alignment checks costs little, but it holds only if planners stop stretching intervals for outages.
Actions taken
- Replace the cracked housing with a redesigned casting that carries a larger fillet radius, and inspect the mating housing for the same crack pattern during that outage.
- Cap the nip load on heavy grades at the level the original housing was rated for, and log every exception so planning can see what a heavier order costs.
- Restore the bearing lubrication interval to the manufacturer's figure, and move alignment checks into a fixed weekly routine that needs a signed waiver to defer.
- Fit accelerometers to both press roll housings and set alarm thresholds from the first month of clean running data, so a rising trend triggers inspection before a crack forms.
Results
The redesigned housing and load cap ran through the next three planned outages without a bearing housing failure. Unplanned stops in the press section dropped from three in the prior period to none in the following period, and the mill recovered most of the production it had lost. Heavy grade orders now run at the lower nip load, with slightly wider reel moisture variation accepted by quality after trial review. The accelerometers flagged a rising trend on the upper roll about six weeks before it would have forced a stop, and the inspection found early wear that was corrected on schedule.
Questions for discussion
- Which operating record would you check first to decide whether a cracked part points to a design problem or a maintenance problem?
- How would you weigh a load cap on a profitable customer grade against the cost of another unplanned stop?
- What evidence would convince you that a vibration alarm threshold is set correctly for a machine you have just commissioned?
Case Study 2
Gluer bottleneck drives defects on a carton line
Harrowgate Carton Company, a folding carton plant that converts coated board for food and personal care customers.
The situation
Harrowgate Carton runs one main line from printed board to palletized cartons, and the folder-gluer at the end of that line sets the pace for the whole plant. Over the quarter, its output fell about 18 percent below the schedule planning had set. At the same time, the stacker reject rate climbed from under 2 percent to about 5 percent. Operators had been running the line faster to cover the shortfall, which opened glue seams on lighter board and skewed flaps on heavy grades. A new customer order with tighter glue placement arrived in the second month and made the problem impossible to ignore.
The challenge
The plant ran two shifts with no spare folder-gluer. Management wanted more throughput without buying a second machine, and the quality manager wanted rejects back under 2 percent before the customer audit. The team had to choose which variation sources to attack first, define acceptance criteria that the shop floor and the customer would both accept, and decide whether faster running was helping the schedule at all.
Analysis
The process map ran from press to stacker, and each station had its own cycle time and tolerance. The folder-gluer was the constraint, averaging about 37 cycles per minute, while the die cutter carried roughly 15 percent slack. Shift logs showed the constraint losing time in three places: changeovers, glue pot refills, and feed jams. Glue bead tolerance was narrow, with bead offset allowed to move only a small fraction of the flap width before a carton failed the fold test. Maintenance records showed nozzle cleaning on a fixed interval rather than on condition, so bead position drifted between cleanings.
Variation on this line sorts into what the line creates and what it inherits from upstream. Board caliper changed by supplier batch, and printed panels carried ink density shifts that altered how flaps folded. Glue defects clustered after shift changes, which pointed to the nozzle routine and to a glue temperature setpoint that operators adjusted by hand. Speed was not the root issue, but it magnified every other source. A supplier caliper check, a glue temperature control loop, and condition-based nozzle cleaning would address most of the loss, while a faster machine setting alone would only raise rejects per hour.
- 1Printing pressInk density shifts with panel coverage and changes flap folding.
- 2Die cutterCarries about 15 percent slack and rarely sets the pace.
- 3Folder-gluerConstraint at about 37 cycles per minute with glue bead drift.
- 4Stacker reject checkCounts rejects at each shift end, the first point of visible loss.
- 5PalletizingDownstream of the reject point and not a source of loss.
Actions taken
- Set a supplier caliper check at receiving, with a pass band agreed on sampled stacks, so batch changes are caught before they reach the press.
- Install a glue temperature control loop on the melter, tie the setpoint to the job recipe, and remove manual adjustment during shifts.
- Move nozzle cleaning from a fixed weekly interval to a trigger based on bead offset readings taken at the start of each run.
- Hold line speed where rejects stay under 2 percent for two consecutive weeks, then raise speed in small steps under the same acceptance check.
Results
Within the third week, the supplier caliper check caught two batches outside the pass band before they reached the press. Glue temperature stability improved once the control loop replaced hand adjustments, and nozzle cleaning moved to condition triggers, which cut mid-shift interventions close to zero. The reject rate at the stacker settled near 1.6 percent and held there for six consecutive weeks. Throughput recovered to within a few percent of schedule, and later speed increases followed the acceptance rule without any return of glue defects.
Questions for discussion
- Which measurement would you place on the folder-gluer first, and how might that choice mislead you if the die cutter has hidden slack?
- How would you set a glue bead offset criterion that both the shop floor and the customer audit would accept?
- When schedule pressure appears partway through a quarter, what evidence should move a plant from a speed increase back to a process fix?
Case Study 3
Waste baler guarding review before full production restart
Ardent Fiber Recovery, a recycling yard that bales corrugated and mixed paper for mill customers.
The situation
Ardent Fiber Recovery added a second horizontal baler at its receiving yard to handle growing volumes of mixed paper and corrugated. The machine feeds through a hopper with a light curtain, closes with a hydraulic ram, and has a rear access panel for maintenance. Commissioning trials ran for the second month, but the safety lead asked for a formal review before the line moved to full volume. Design documents showed the interlocks as installed, yet operating data from trials logged two near-misses at the feed opening during jams, and crews had been clearing jams with the ram partly open.
The challenge
The project manager wanted the baler at full volume before the end of the quarter, because a mill contract depended on the extra tonnage. The review had to separate hazards controlled by engineering from those that depended on procedure, and identify missing documents before the approval gate. Maintenance records also showed the hydraulic hose inspection signed off without a recorded pressure test, which raised a separate question about stored energy in the system.
Analysis
The hazard inventory started with the ram closing force, the feed opening, the rear access panel, and the staging area where operators pull bales. These carry crush, entanglement, and stored energy risks. Some controls were straightforward. The light curtain and rear interlock can be verified with a functional test and logged. Others were weaker than they appeared. A light curtain protects the feed opening only when its muting logic matches the feed cycle, and the commissioning data showed the muting window set wider than the vendor default. Quality tolerances on bale density are not safety measures on their own, but overweight bales were jamming the ram and prompting the unsafe clearing practice.
Machine guarding, energy isolation, and hydraulic requirements apply to this installation, and the review mapped each control to the requirement it was meant to satisfy. The gaps were documentary as much as physical. The risk assessment had not been updated after the feed door was modified, the lockout procedure covered the main disconnect but not the accumulator, and no sign-off existed for the muting configuration. The hose inspection was logged, but the pressure test the hose supplier specifies for that service was missing. The approval gate should stay closed until the assessment, the procedure, and the test record are current and signed by the people responsible for each.
- Early in trialsTwo near-misses logged at the feed opening during jam clearing.
- Start of reviewSafety lead requests formal review before full volume production.
- Second week of reviewGaps logged across assessment, procedure, and test records.
- Two weeks laterMuting reset, procedure revised, and hose test completed on schedule.
- About five weeks inSign-off issued once records matched the installed machine.
Actions taken
- Reset the light curtain muting window to the vendor default, then validate it with a documented test that confirms the feed opening stops the cycle in every position.
- Revise the risk assessment for the modified feed door, and add a stored energy step to the lockout procedure covering accumulator bleed-down.
- Stop jam clearing with the ram partly open, and add an upstream bale density check so overweight bales do not reach the chamber.
- Complete the hydraulic hose pressure test, file the record, and require sign-off from maintenance and safety before the approval gate is passed.
Results
The approval gate passed about five weeks after the review opened, later than the project manager had hoped but ahead of the contract date that required full volume. Re-validation showed the feed opening was protected in every tested cycle, and two crews practiced the stored energy step before the revised procedure went live. Jam clearing events fell sharply after the density check was added, and no further near-misses were logged after restart. The hose test found one line due for replacement, which was swapped before the machine reached full throughput.
Questions for discussion
- Which of the gaps you found would you treat as a design issue and which as a documentation issue, and how would that change who owns the fix?
- What evidence would you need before accepting a muting configuration that differs from the vendor default?
- How would you handle a schedule conversation when the approval gate is the only remaining constraint?
Case Study 4
Recovery choice after a gearbox casting rework
Brindle Paper Converters, a converter of specialty paper and printed labels with two sheeting lines and one flexographic press under installation.
The situation
Brindle Paper Converters ordered a new eight-color flexographic press to add capacity for printed label and carton work. The contract called for delivery early in the second quarter, with a short install window ahead of a customer launch. The press vendor sourced its main gearbox from a subcontractor, while the foundation was designed by the building contractor on a separate schedule. In the third month, the gearbox casting arrived out of dimensional tolerance and needed rework, and the foundation drawings were revised twice after the frame layout changed.
The challenge
The milestone slipped about nine weeks, and the customer launch could not move. Finance had approved a budget built on an on-time install, and the overrun already stood at roughly 210,000 dollars in expedite fees, idle crew time, and temporary tooling. The steering group had three recovery options, each with a different cost and risk profile, and it wanted a clear comparison before committing.
Analysis
The schedule problem was three problems stacked together. The supply issue was the gearbox casting, which the subcontractor had queued behind its own customers. The design issue was the foundation, drawn to an earlier frame layout and needing a new anchor pattern. The coordination issue was that the building contractor and press vendor reviewed drawings in separate meetings, so the revised layout reached the foundation team late. Site constraints were tight: the install crane was booked for another line, and the press could not be placed until the floor cured and the anchors were set. The vendor's commissioning plan needed about four weeks at the stated run rate, which left very little float.
Recovery plans should be compared on the same basis, so each option was reduced to the cost of earlier delivery, the risk of a second slip, and the effect on maintenance and print quality. Option one ran the gearbox rework on a premium schedule and kept the current foundation drawings, which lowered schedule risk but left the layout untouched. Option two moved gearbox work to a second subcontractor and accepted a shorter warranty period. Option three staged the install so two of the eight units could commission first while the remaining foundation section cured. Each option moved the launch date differently, but they also added different numbers of new interfaces, and new interfaces are usually where a second slip begins.
Actions taken
- Approve the premium gearbox rework with a hold point, so the first corrected casting is dimension checked before any further machining starts.
- Issue one combined drawing set that the building contractor and press vendor sign together, and freeze the anchor pattern before any concrete is poured.
- Stage commissioning so the first two units run trial jobs while the remaining foundation section cures, and move the first customer orders to those units.
- Set a weekly joint review with both suppliers and the launch customer, and track float against the commissioning phase rather than the install date.
Results
The gearbox rework was approved, and the first corrected casting passed dimension checks after one further adjustment. The team staged commissioning, and two units began trial jobs in the third week after the revised milestone, which allowed the first product lines to launch on the committed customer date. The remaining units followed as the foundation cured. Total overrun cost came in below the premium estimate because the staged plan avoided a second expedite cycle, although the joint review added meeting time for both vendors. Final commissioning ran with about two weeks of float, which the team judged too thin to repeat.
Questions for discussion
- Which of the three recovery options would you reject first, and what evidence would change your mind?
- How would you structure one drawing approval process across a building contractor and an equipment vendor?
- When float is thin, how would you compare the cost of an extra vendor review meeting with an expedite fee?
Case Study 5
Cutting steam use on a containerboard dryer section
Larkspur Board Mill, a recycled containerboard mill with one machine and an in-house steam plant.
The situation
Larkspur Board Mill runs a single machine that makes recycled containerboard at roughly 1,300 tons per week. Steam for the dryer section accounts for most of the mill's energy use, and the steam plant has run close to its firing limit during cold weeks. Management set a target to cut steam per ton by about 8 percent without reducing machine speed or board strength. The options under discussion were hood exhaust heat recovery, a condensate flash upgrade, and a lower sheet moisture target at the press. Each touched a different part of the process, and none could be tested on the live machine without a planned outage.
The challenge
The sustainability lead wanted a figure to report to customers, while procurement asked whether any change would shift fiber losses or waste. Engineering had to compare the options by energy saved, capital cost, maintenance burden, and effects on the board tolerances that customers audit. The dryer section was not metered separately, and the existing data was patchy, so the comparison needed new measurements before any number could be trusted.
Analysis
The comparison started with operating data. Steam use per ton was tracked for the whole plant, but the dryer section had no dedicated meter, so the team installed temporary meters on the hood supply and condensate return and ran them through a full production cycle. The data showed a large share of energy leaving through the hood exhaust as sensible heat, and condensate returning hotter than the flash tank could use. Quality tolerances set the limits. Caliper and reel moisture are tightly specified, and moisture drives box compression, so any change to the moisture target had to be checked against print and compression results. Maintenance records showed the hood exhaust fan had been bypassed twice during shutdowns, which reduced its reliability as a recovery point.
Lifecycle and operational measures have to sit side by side. A heat recovery exchanger on the hood exhaust cuts purchased fuel directly, and the mill's own cost model suggested a payback within a few years, but it adds fouling and corrosion duty that maintenance must schedule. A condensate flash upgrade costs less and has fewer moving parts, but it recovers less energy. Lowering the press moisture target saves steam with no capital at all, yet it can push reel moisture out of tolerance and affect board stiffness. The measures that matter are steam per ton, caliper and moisture variation, fan availability, and waste from rejected rolls. An option that cuts steam while raising rejects has not improved the mill.
- Hood recovery Captures exhaust heat but adds fouling and corrosion duty.
- Flash upgrade Lower capital and fewer moving parts, with a smaller energy return.
- Moisture target Saves steam directly but narrows the reel tolerance margin.
- Fan reliability Keeps the recovery point available through every production cycle.
Actions taken
- Install heat recovery on the hood exhaust with a fixed cleaning interval, and check fouling at each scheduled outage during the first months of operation.
- Add the condensate flash upgrade at the same outage so the hotter return is used rather than vented, starting with a flash tank sizing check.
- Lower the press moisture target in small steps, checking reel moisture, caliper, and box compression at each step before the next adjustment.
- Return the hood exhaust fan to continuous service, require a signed work order for any bypass, and track fan availability alongside steam per ton.
Results
Steam per ton fell by about 9 percent over the following quarter, slightly above target, with no change to machine speed. Hood heat recovery delivered most of the reduction, and the condensate flash upgrade added a smaller share. The moisture target was reduced in narrow steps, and reel moisture stayed inside the customer band in every sampled roll. Box compression results held within the earlier range, and the rejected roll rate did not rise. Fan availability improved, although the first fouling check found more deposit than predicted, so the cleaning interval was shortened.
Questions for discussion
- Which measure would you rank first when heat recovery and a moisture target change are compared, and why?
- How would you decide whether a fouling check interval on a hood exhaust exchanger is set too long?
- What would you need to see in the waste and reject data before you accepted a steam saving as a real improvement?
Learning Resources
Industrial CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.