Industrial Fan System Effects

Six Real-World Case Studies

At AMCA International's conference, our team presented six case studies on advanced system effects. The cases came from real installations across manufacturing, wastewater treatment, correctional facilities, and pulp-and-paper production. The pattern held across all six. The engineering looked sound on paper, but unaccounted system effects drove up performance and maintenance costs in the field.

A fan is only as good as the system around it. If the design ignores inlet conditions, duct layouts, dampers, or commissioning practices, even a well-built, properly specified fan can vibrate, overheat, or fail early. Here are six examples that show exactly how that happens, and what to do about it.

Manufacturing

Case 1: Commissioning Before the System Was Complete

A large manufacturing facility ran a general exhaust system built around multiple centrifugal fans. A VFD drove each one, and a control sequence held constant pressure inside a shared header. Dampers in each branch controlled the airflow rates. During commissioning, the team reported inadequate airflow and elevated vibration.

The cause came down to timing. Construction schedules forced the team to commission the fans before crews finished the full system. To compensate, the team partially blocked the main headers and added smaller dampers to create "false loads." Header pressure read correctly, but airflow stayed low. The first two or three fans had enough airflow. The last one or two ran starved of air, near stall.

The lesson: pressure readings alone don't tell you what a fan is actually doing. A fan in stall can produce the right static pressure while moving the wrong amount of air. The wear it builds up stays invisible until something fails.

Wastewater Treatment

Case 2: 75% of Airflow Through a Single Quadrant

A wastewater odor-control application used two Arrangement 8 centrifugal fans in parallel, one running and one on standby. Each fan handled 30,000 cfm at 18 inches w.c. Within the first six months, the running fan developed oscillating vibration. The vibration grew severe enough to trip the high-vibration switch on a regular basis. Eventually, Fan 1 failed outright.

CFD analysis revealed the root cause. The duct split immediately upstream of the fans, so 75% of Fan 1's airflow entered through a single quadrant of the wheel. Fan 2 had a similar issue, though slightly less severe. The oscillating vibration that resulted caused shaft turndown at the bearing seat.

The air permit made things worse. It required exact airflow over a 24-hour rolling average, so the owner programmed the VFDs for constant real-time speed adjustments. The fans never settled into a stable operating point. The fixes came in three parts. The team added P-traps to handle water buildup in the duct. They rewrote the air permit to drop the real-time speed corrections. And they added targeted Unistrut reinforcement to cut air-induced vibration.

The lesson: when air enters a fan wheel asymmetrically, it creates vibration the wheel can't handle. Straight duct between elbows, tees, and fan inlets isn't a nice-to-have. It's what protects the bearings and shaft.

Correctional Facilities

Case 3: A 24.5% Free Area Loss Nobody Calculated

A correctional facility with roughly 200 cells ran a ventilation system that couldn't deliver the required airflow to each cell. Every cell had a 6-by-6-inch ventilation opening. Per code, each opening also carried a 0.75-inch-diameter hardened steel security bar running both horizontally and vertically across it.

We compared the as-built drawings to the design drawings. Then we checked the AMCA-certified fan performance data against the on-site test-and-balance reports. The picture was clear: the system effect calculation left out the security bars. Across 200 openings, those bars reduced free area by 24.5%. The original fan ran near its maximum RPM, so it had no headroom to compensate.

The facility ultimately replaced the fan with a larger one sized for the actual losses.

The lesson: every component in the airflow path counts, even the ones added for non-mechanical reasons. When a system has repeated features (200 of anything), small per-unit losses compound quickly. And when a fan runs near its operating limits, it has no margin to absorb the surprises.

Pulp and Paper

Case 4: When a Closed Damper Cooks a Fan

A pulp-and-paper process exhaust application used a 316 stainless steel centrifugal fan designed for ambient air, roughly 75°F. After three months of operation, vibration exceeded 0.8 inches per second. When the team inspected the fan, they found the epoxy coating cooked off the housing.

Maintenance records explained what happened. During an emergency breakdown elsewhere in the plant, crews held a butterfly damper closed for six to eight hours. That damper sat directly upstream of the fan. The fan kept running with its inlet blocked. Air recirculating inside the housing generated enough heat to push temperatures over 500°F. That heat warped the wheel and destroyed the coating.

We redesigned the wheel with high-strength stainless components, including Inconel in critical structural areas. For other facilities running similar applications, we raised the maximum operating temperature spec to 550°F. That gives the design enough margin to survive a similar event.

The lesson: a fan that can't move air can't cool itself. Operating procedures and mechanical design have to account for what happens when a fan keeps running into a closed system. Eventually, it will happen.

Poor Layouts Across Applications

Case 5: When Footprint Constraints Override Good Design

Two installations show what happens when there isn't enough space to do it right.

In the first, a tee and a control damper sat directly upstream of a centrifugal fan. No straight duct ran between them. The customer reported a loud, harsh-sounding fan. After analysis, the customer adjusted the damper to a 45-degree angle to try to manage the airflow. Within two weeks, the turbulent inlet conditions had cracked welds on the centrifugal wheel. We eventually recommended a VFD to control airflow instead of dampering it that close to the inlet.

The second was more extreme. To fit the fan into a tight footprint, the contractor installed a hard inlet bend of roughly 135 to 180 degrees. Then they put an inlet box on the fan outlet. The fan ran with severe vibration. A different arrangement, specifically a square design rather than Arrangement 01, would have eliminated the system effects entirely.

The lesson: when footprint forces compromises near the fan inlet or outlet, change the fan arrangement. Don't bend the duct harder. Talk to the manufacturer before the duct routing is final.

Complex Wastewater Systems

Case 6: 18 Months of Staged Startups Without Air Balancing

This system was complex from the start: six supply fans, twelve area exhaust fans, and three odor-control fans. They operated in some combination of series and parallel, feeding a common odor-control duct from six branches. The target was 91,000 cfm overall. The fans were belt-driven centrifugals, with the odor-control units on VFDs.

The factory balanced and AMCA-tested the fans before shipment. But the system came online in stages over 18 months. Each time a new fan went in, the contractor declined to balance the system, preferring to wait until everything was in place. In hindsight, the result was predictable: erratic airflow, extremely noisy operation, and frequent vibration trips. The oscillating vibration grew severe enough to cause shaft turndown on the odor-control fans.

Once the crew finally air-balanced the full system, the vibration issues resolved. The team replaced the bearings on the odor-control fans with concentric bearings to handle any residual oscillation. They also replaced two shafts that the turndown damage had already ruined.

The lesson: air balancing isn't a final step. It has to happen every time you bring a fan online in a connected system. System pulses are a strong sign that the air isn't balanced. Ignore them, and those pulses do real mechanical damage.

What Ties These Six Cases Together

Every one of these installations used a properly built, properly rated fan. AMCA certification confirms a fan's performance in lab conditions. But lab conditions don't include elbows two duct diameters from the inlet, security bars across 200 openings, or a damper held closed for six hours during a plant emergency. Those are system effects. They fall on the system designer, the installer, and the commissioning team, not the fan itself.

Three practices prevent most of what we saw in these six cases:

• Account for system effects during design. AMCA Publication 201 and the Fan and Air System Applications Handbook are good starting points. Every elbow, tee, damper, and obstruction near the fan inlet has a cost.

• Maintain straight duct between components. Air entering a fan wheel needs to stay as laminar and symmetrical as possible. The straight-duct rules in AMCA's recommendations aren't arbitrary. They keep the bearings and shaft alive.

• Air-balance every time you bring a fan online. Not just at final commissioning. Every stage.

If you're designing a system or troubleshooting one that's already in trouble, our application engineers can help. They've walked installations like these and can spot the issues before they become failures. For related reading, see our overview of AMCA Certified Fans, which explains why third-party performance verification matters. Our guide to centrifugal fan selection covers how CFM and static pressure decisions feed into the issues above. And our piece on choosing the right industrial fan shows how installation conditions often look like fan failures.

Phone: 1-800-336-3267

Email: info@hartzell.com

About the Presentation

Our team presented these six case studies at AMCA International's 2026 conference, in the session "Advanced System Effects: Causes and Remedies." Three members of the Hartzell Air Movement team delivered the presentation: Ryan Cassel, Fiberglass Product and Market Manager; Adam Habegger, Metal Centrifugal Product Solution Engineer; and Justin Meyer, Sales Application Engineer. Together, they bring more than 50 years of combined experience across industrial fan applications, pollution-control equipment, and corrosion-resistant materials.

Frequently Asked Questions

What is a fan system effect?

A system effect is any performance loss that the installation conditions create, rather than the fan itself. Elbows, tees, dampers, security bars, transitions, and other obstructions near the fan inlet or outlet disrupt the laminar airflow the fan needs. The result: reduced airflow, higher static pressure loss, vibration, noise, and faster wear. All of this happens even when the fan runs exactly as rated in lab conditions.

If my fan is AMCA certified, why am I still having performance problems?

AMCA certification verifies that a fan performs as rated under controlled lab conditions. It doesn't account for the conditions of the installation itself. An elbow two duct diameters from the inlet, a damper held closed during a process upset, 200 security bars in a corrections facility: these are all system effects. They fall on the designer, installer, and commissioning team. A properly certified fan can still underperform if no one designs and installs the surrounding system with those effects in mind.

How much straight duct should I leave between a fan and the nearest fitting?

The general guidance from AMCA Publication 201 calls for at least three duct diameters of straight duct upstream of the fan inlet, and two to three diameters downstream of the outlet. Elbows, tees, and dampers call for more. The exact requirement depends on the fan type, the airflow rate, and the specific fitting. When in doubt, get the manufacturer's input before you finalize the duct routing.

Why does a closed damper damage a running fan?

When a fan keeps running into a closed system, air recirculates inside the housing instead of moving through it. That recirculation generates heat. Housing temperatures can climb above 500°F within hours, which can warp the wheel, destroy coatings, and damage bearings. Fan systems should include controls or interlocks that prevent extended dead-head operation. The mechanical design should also carry enough thermal margin to survive an unplanned closure.

When should you air-balance a system?

Every time you bring a fan online in a connected system, not just at final commissioning. Multi-stage projects that defer balancing until the end produce the exact vibration, noise, and shaft-turndown issues from Case 6. System pulses, oscillating vibration, and erratic airflow all signal an unbalanced system. The mechanical damage from running that way can be significant.

When should I involve a fan manufacturer in the design process?

As early as possible, and certainly before anyone finalizes the duct routing. Most of the cases above stem from decisions that came after someone specified the fan but before the installation was complete. Think footprint changes, late-breaking obstructions, staged commissioning, or modifications to fit the available space. A manufacturer's application engineer can flag system effects and recommend an arrangement that fits the layout. That helps avoid the kind of rework that shows up six months into operation.