Environmental and Sustainability Engineering Case Studies
Technical5 Case StudiesUpdated Sep 2026
Environmental and sustainability work rarely fails in the abstract. It fails at a pump seal, a cooling loop, a permit condition nobody reread, or a supplier shipment that arrived three weeks late. The five cases here come from plants, utilities, and project teams where the problem was specific and the documentation was mostly sound. Each case follows the engineer's view: what the operating data showed, which constraints were fixed and which were negotiable, and which corrective actions held up. Most of the answers were already in the maintenance logs and specifications before anyone asked the question.
Case Study 1
Membrane cassette failure in a wastewater reclamation plant
Ostrander Creek Water Reclamation, a regional utility serving about 180,000 residents through two membrane bioreactor trains.
The situation
The Ostrander Creek plant runs two membrane bioreactor trains, each sized for a design flow of 14,000 cubic meters per day. In the warmest weeks of the summer, Train B began passing solids into permeate, and turbidity readings climbed above the 0.15 NTU limit on several afternoons running. Operators first suspected the cassettes, which were the newest hardware on site. Early inspection showed fiber breakage concentrated near the header potting rather than spread randomly across the module. The membranes had been in service for about five years, within the vendor's stated life, yet the failure rate was well ahead of what the original design basis assumed.
The challenge
The open question was whether the failure came from the membrane material, the operating regime, or how the cassettes were installed and cleaned. Each explanation pointed toward a different fix and a different price. A material change meant replacing every cassette, a capital request well beyond the plant's annual budget. A process change meant adjusting aeration and cleaning frequency, which the plant had never validated at this load. Maintenance records showed the chemical cleaning interval had been stretched twice to save reagent, with no one tracking the cumulative exposure.
Analysis
Teams grounded in Environmental and Sustainability Engineering usually start by separating what the design basis promised from what the plant has actually asked of the equipment. In this case the design basis assumed a transmembrane pressure ceiling and a chemical exposure budget per year. Operating data, pulled from the historian over the full service life, showed transmembrane pressure running near the ceiling for longer stretches than modeled, and the cleaning log showed reagent dosing had drifted lower. The permit set strict limits on permeate turbidity, so a few broken fibers could breach compliance even when the average looked fine. The data pointed to exposure, not defect, as the leading cause.
The maintenance records added a second layer. Cassette integrity tests had been performed at the scheduled interval, but the pass threshold had been relaxed in an earlier revision of the operating manual, so modules that should have been pulled were kept in service. Teardown of three failed modules showed the breaks clustered at the potting interface, pointing to repeated flexing during backwash. The corrective options were a tighter integrity threshold with more frequent testing, a revised cleaning recipe with restored reagent dosing, and a header redesign. The team ranked them by how directly each one addressed the observed mechanism, not the most visible symptom.
Actions taken
- Restore the original integrity test threshold and test every module at the interval set in the design basis, not the revised manual value. Pull any cassette that fails on the first attempt.
- Reinstate the full chemical cleaning recipe and log every dosing change with the operator's initials and the reason. Reagent savings were real but small compared with the cost of premature module replacement.
- Add a header support bracket and reroute backwash suction so the potting interface no longer carries repeated flexing. The bracket was specified against the vendor's load case and checked by the structural reviewer.
- Set a quarterly review of transmembrane pressure trends against the design ceiling, with a trigger that opens a root-cause file before any permit limit is reached.
Results
Within the first quarter after the revised test regime, Train B failed two more modules, both at the potting interface, which confirmed the mechanism. After the bracket and reagent changes, permeate turbidity stayed below the permit limit across the next full summer, and the monthly replacement rate dropped to a level the operations budget could carry. The replacement program cost a fraction of the material-change proposal. Two residual concerns remain: the cassette supplier has not yet published fatigue data for the new bracket geometry, and the team has asked for a second teardown after the next cleaning cycle.
Questions for discussion
- Which evidence would have justified a material change before the process data was pulled?
- How should a plant decide when a relaxed inspection threshold is a reasonable operating choice and when it is a hidden design change?
- What monitoring signal would have revealed the flexing mechanism before any permit limit was approached?
Case Study 2
Recovering a polymer line from throughput and defect loss
Ferncliff Recycled Polymers, a pellet producer running one wash and extrusion line across two shifts.
The situation
The plant's wash and extrusion line was designed for a nameplate output of 2,400 kilograms per hour of pellet. Over the course of the spring, output settled near 1,750 kilograms per hour while the rejected share of batches rose from under two percent to nearly seven percent. Customers flagged yellowing and gel specks, and two lots failed melt flow index testing. Production leads initially blamed the feedstock, since bale quality had shifted with the change in municipal collection contracts. The line itself had also been modified twice in the previous year to add a second screen changer, and nobody had mapped how those changes affected dwell time.
The challenge
The team had to decide where to start. Throughput loss and defect rise could come from the same cause, or from two separate ones, and the answer changed where the engineering effort should go. Melt flow variation was the obvious quality signal, but it was also the hardest to link to a single process step. Management wanted recovery within the quarter, and the line could not stop for long trials without losing committed shipments.
Analysis
The first step was a process map drawn from the line's actual flow, not the original drawing. Residence time in the wash tanks was set by design at a fixed value, but the historian showed the flake level and feed rate drifting enough to change contact time by a measurable margin during the shift. Quality tolerances on color and melt flow index were tight, and rejects clustered after the second screen changer was installed, when screen swap intervals lengthened. Plotting rejects against screen pressure and wash temperature produced a clearer picture than any single average. That turned a vague complaint into testable causes.
Variation came from three places: bale quality, tank residence and screen loading, and maintenance timing. The maintenance log showed screens were changed by calendar, not by pressure rise, so some ran past useful life while others came out early. Design constraints capped extruder throughput, so the quickest gain was more consistent screen timing, not a bigger machine. The team ranked actions by expected effect on rejects, then by how quickly each could be trialed without stopping the line. Acceptance criteria were set before any change: rejects under three percent, output at least 2,100 kilograms per hour, and melt flow in band for ninety-five percent of lots.
- Measure dwell Log residence time, screen pressure, and reject timing for every lot across two shifts.
- Isolate variation Separate bale quality effects from process and maintenance effects using matched shift comparisons.
- Trial one change Run one change on one shift, holding all other settings at baseline.
- Verify and standardize Confirm the gain against acceptance criteria, then update the work instructions and maintenance schedule.
Actions taken
- Move screen changes from a calendar interval to a pressure-differential trigger, and log the pressure at each change so the trigger can be tuned over several weeks.
- Hold tank residence within a fixed band by tying feed rate to flake level, with a check at the start of each shift against the lot record.
- Run a two-week trial on one shift with the new screen rule and the residence band, compared against the untreated shift running the same bale lots.
- Bring the acceptance criteria to the weekly review and stop the change if rejects rise above the baseline for two consecutive lots, since a slower line that ships clean product is preferable.
Results
The shift trial ran for fourteen production days. Reject rate on the trial shift fell from roughly six percent to just under three percent, and output rose to about 2,050 kilograms per hour, still short of target but well above the prior average. Melt flow index stayed inside the customer band for ninety-six percent of lots. The result missed the criteria, so the trial ran one more week with a tighter residence band. Output reached the 2,100 target on most days after that, and the remaining gap traced to a pelletizer knife set on a longer service cycle.
Questions for discussion
- Which measurement would have shown the screen timing problem earliest, and how often should it be taken?
- When a process change lowers throughput, what evidence should be required before the line returns to its prior rate?
- How should acceptance criteria be weighted when throughput and quality targets pull in opposite directions?
Case Study 3
Digester permit review with missing hazard evidence
Greywater Hollow Dairy Cooperative, a group of member farms proposing a shared anaerobic digester for manure and food residue.
The situation
The cooperative wanted a shared digester for about 60 tonnes per day of manure and food residue, with biogas feeding a heat and power unit. The county asked for a safety and permitting review before the grading permit would issue. The package held a process flow diagram, a site layout, and a vendor data sheet. The vendor sheet listed hydrogen sulfide as a potential constituent but gave no removal or monitoring basis. The package had no hazard analysis, gas detection plan, or documented pressure relief sizing. Two member farms had already scheduled earthwork for the following season.
The challenge
The central concern was that the hazards were known to the designers but not written down in a form the county or the workers could use. The digester would run as a shared asset, which meant decisions about gas handling, confined entry, and emergency response would fall to members who had never run anaerobic equipment. Approval gates were unclear. The grading permit was tied to the design review, but the design review was tied to a hazard analysis that did not exist, and the schedule left little room for sequencing.
Analysis
The digester volume, the gas storage pressure, and the proximity of the combined heat and power unit to occupied buildings were all fixed in the preliminary layout, so hazards had to be assessed against that geometry rather than an ideal site. Operating data from two comparable farm digesters was useful but limited, since their feedstock mix was lower in sulfur. The sulfide concentration expected in this biogas was therefore an estimate, not a measurement, and it governed everything downstream: corrosion allowances, scrubber sizing, and the worker exposure limits for entry. Hazard evidence had to show both the expected range and the margin above it.
Maintenance records from the two reference digesters were the most useful evidence, because they showed what had actually failed. Both had gas detector calibration logs that lapsed for several weeks at a time, and one had a relief valve tested only at commissioning. The local fire code, the state air permit, and the adopted gas piping standard each set documentation and inspection requirements that the package did not yet meet. The review ranked gaps by severity: missing gas detection and relief sizing first, then the confined entry procedure, then the odor control plan. Earthwork waited on an accepted hazard analysis, and gas admission waited on filed commissioning evidence.
Actions taken
- Commission a hazard analysis covering hydrogen sulfide, biogas pressure, and confined entry, with the sulfide estimate replaced by a measured range from a pilot sample.
- Size the pressure relief devices against the worst credible gas generation case and document the calculation, so the county reviewer can check the basis without redoing it.
- Specify a fixed gas detection layout at the digester, the gas storage, and the engine room, with calibration on a fixed schedule and logs kept on site.
- Hold earthwork until the hazard analysis is accepted, even though two farms lose a planting window, and gate gas admission separately on filed commissioning evidence.
Results
The hazard analysis took most of the next quarter and produced a sulfide range about three times the vendor's assumption, which changed the scrubber specification and the entry procedure. The relief sizing revealed that the original device would have been undersized for a stalled feed condition. Two member farms lost their planting window and paid for temporary rental of spreading equipment. The county accepted the revised package, and the grading permit issued in the sixth week after resubmittal. Gas admission remained blocked until commissioning evidence was filed, and the cooperative now carries a monitoring schedule its members must staff.
Questions for discussion
- Which evidence should a reviewer require before accepting an estimated hazard concentration in place of a measured one?
- How should approval gates be sequenced when a delay carries a real cost for the people the project is meant to serve?
- What inspection record from a comparable facility would most change your view of the proposed control layout?
Case Study 4
Reclaimed water skid delayed by housing supply
Pellham Ridge Utilities, a municipal water utility adding reclaimed water treatment for parks and athletic fields.
The situation
Pellham Ridge Utilities planned to bring a reclaimed water filtration skid online before the irrigation season, supplying treated water to parks, a school field, and a golf course under a contract with late delivery penalties. The equipment and installation budget stood at 2,850,000. The pressure vessel housings were the long-lead item, and the supplier notified the team in the eighth week of fabrication that the housing order would arrive eleven weeks late. The skid could not be commissioned without them, and the contract milestone was fixed. The design team had already completed the piping layout around the original vessel dimensions, so any substitute had to fit the same footprint.
The challenge
The team needed to recover at least eight of the eleven weeks without raising the budget beyond its contingency. Three routes were on the table: expedite the original housings by air freight, substitute housings from a domestic fabricator that needed a new hydraulic check, or split commissioning so the skid ran on a reduced train while the full capacity waited. Each route moved a different cost, and none of them could skip the review of the revised hydraulics.
Analysis
The schedule math started with the fixed constraints. The contract milestone sat on a specific irrigation start date, and the reclaimed water quality permit set tolerances on turbidity and log reduction that the substitute housing still had to meet. Operating data from the existing potable plant gave a rough cycle time for the membrane train, which allowed the team to estimate how much capacity a reduced train could deliver during the first weeks of service. That estimate was the basis for comparing the split commissioning route against the other two. Quantifying the trade-offs meant putting a number on each option: freight cost, recheck fee, and expected penalty days.
Maintenance records from the utility's two earlier skids were the best evidence for the split option. Both had run a reduced train during repairs, and their logs showed fouling and cleaning rates that set a baseline. Expedited freight added about 96,000 to the budget and recovered five weeks. The domestic substitute recovered six weeks for about 140,000 plus a hydraulic recheck. Split commissioning cost about 60,000 in operator hours and temporary controls, and it met the milestone at roughly half capacity. Only the split route avoided new hardware risk, but it required the design team to show the reduced train still met the turbidity limit.
Actions taken
- Approve split commissioning so the skid runs on a reduced train for the first irrigation weeks while the full train waits for housings.
- Pay for expedited freight on the housings already in fabrication, capped at the contingency so the full train still returns within the season.
- Submit the revised hydraulic arrangement for the reduced train to the permit reviewer with turbidity and log reduction data before any water leaves the skid.
- Write the contract notice of partial delivery now, with penalty exposure stated per day, instead of waiting for the milestone to be missed.
Results
The team chose split commissioning with expedited freight. The reduced train went live on the milestone date at roughly half capacity, and turbidity stayed inside the permit limit from the first day. The housings arrived five weeks sooner than the supplier's revised date, and the full train came online in the eighth week of the season. Penalty exposure was limited to the days the school field demand exceeded what the reduced train could supply, and the customer accepted a prorated schedule. Total added cost was about 156,000, which the contingency absorbed without a budget amendment.
Questions for discussion
- How should a split commissioning route be compared with full schedule recovery when reduced train performance is still uncertain?
- Which contract terms are worth negotiating early when a supplier delay is still a forecast rather than a confirmed loss?
- What records from earlier projects would you check first to estimate partial capacity at a milestone?
Case Study 5
Cutting gas and film use in a bakery line
Oakhollow Grain Foods, a regional bakery producing about 9,000 loaves per shift on two tunnel ovens.
The situation
Oakhollow's main bake line ran two tunnel ovens, a spiral cooler, and a packaging line that wrapped loaves in polypropylene film. Gas use per thousand loaves had risen steadily as the product mix moved toward denser grain formulations, and the packaging scrap rate sat above the level the packaging supplier had quoted. The company held a commitment to cut energy intensity and packaging waste by a fifth within the next planning cycle, with the same shelf life and crust quality customers had come to expect. Management asked for options that would not require new ovens or a change in the bread recipe.
The challenge
Energy and packaging were tied to each other in ways the original project scope had not anticipated. Lower oven exhaust temperature might cut gas use but could extend cooling and raise moisture loss, which shifts the shelf life. A thinner film would cut material, but it could increase seal failures and returns. Each option needed a lifecycle view that included film production and disposal, and an operational view that protected loaf quality and line speed.
Analysis
Start with the baseline the plant could defend. Gas meter readings were logged per oven zone per batch, which allowed the team to separate preheat losses from bake energy. Scrap records came from the packaging line and were tagged by cause, so film edge trim, seal rejects, and misprints could be counted separately. Lifecycle measures were kept simple: energy per thousand loaves, film mass per thousand loaves, and emissions from gas combustion. Each option was tested against crust color, moisture at dispatch, and seal integrity at line speed. A change that cut energy but moved any of those outside tolerance was rejected before costing.
Four options were compared. Heat recovery from oven exhaust to preheat incoming air cut gas use by a measured share, with a capital cost but no change to the bake profile. Lowering exhaust temperature saved more gas on paper but lengthened cooling and raised moisture loss, which pushed some loaves outside the dispatch tolerance. Film downgauge reduced material per thousand loaves, but the seal bar maintenance records showed rising rejects with each thinner trial, so it was restricted to a narrow SKU set. Seal bar replacement on a fixed schedule cut misprints without any capital spend. The ranking screened out options that threatened quality, regardless of lifecycle numbers.
- 1Baseline meters and scrapLog gas per zone and tag every packaging reject by cause before any change.
- 2Screen against tolerancesReject any option that moves crust color, moisture, or seal strength outside limits.
- 3Pilot the top optionsRun heat recovery on one oven and a film trial on two SKUs.
- 4Commit by lifecycle scoreScale only options whose gas, material, and emissions gains hold up over the pilot.
Actions taken
- Pilot heat recovery on one oven with metered gas readings for a full month, and commit the remaining capital only if the saving holds.
- Replace seal bars on the fixed schedule in the maintenance records and log misprint counts by cause, so film savings are judged against actual rejects.
- Trial a 12 percent film downgauge on two high-volume SKUs only, with seal strength tested at line speed on every shift for two weeks.
- Keep the bake profile unchanged and do not lower exhaust temperature, since the moisture loss and shelf life cost outweighs the gas saving.
Results
The heat recovery pilot cut gas use per thousand loaves by a little over a tenth in the first month, and the full installation was approved after the pilot readings held. Seal bar replacement on schedule reduced misprints by roughly a third, which lowered film waste even before any downgauge. The two-SKU film trial produced no seal failures at line speed, and the team expanded it to a third SKU in the next cycle. Crust color, moisture at dispatch, and shelf life stayed inside their tolerances throughout. The energy and packaging targets for the planning cycle were met.
Questions for discussion
- Which lifecycle measure should carry the most weight when a film saving and an energy saving point in different directions?
- How long should a pilot run before its readings are trusted enough to justify the remaining capital?
- What operational signal would show that a quality limit is being approached before a customer complaint arrives?
Learning Resources
Technical 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.