Emergency Sulzer Repair: A 7-Step Checklist for Pumps, Compressors, and the RTA96-C

If you are reading this because a Sulzer pump just tripped offline, or the control system on your Wärtsilä-Sulzer RTA96-C dizel motor is doing something strange, you need the condensed version. This is the checklist I use when there is less than 48 hours before a deadline.

I am a rotating equipment specialist at a Gulf Coast service company. I have coordinated 40+ emergency repairs in eight years, including a same-day turnaround for a chemical plant in Corpus Christi. Missing that deadline would have triggered a $50,000 penalty clause. This is not textbook theory. It is what got the plant back online.

Use this checklist if you have a genuine failure, not just a strange reading; less than 48 hours to make a decision; and a specific part, repair, or replacement needed. If you are still in the "we should study this" phase, stop. This is not your checklist.

Here are the seven steps, in order.

Step 1: Hold a Five-Minute First Congress

The first congress of stakeholders—maintenance, operations, procurement, and the vendor who answers the phone—should happen immediately. Not a formal meeting. A five-minute huddle around the panel.

Here is the trap: everyone thinks "emergency" means the same thing. I once said "as soon as possible" to a supplier. They heard "whenever convenient." Result: delivery two weeks later than I expected. The client's project stalled because I did not define urgency in writing.

At that first huddle, write down:

• What exactly failed?
• What is the deadline, and what depends on it?
• Who has the authority to approve overtime, air freight, and vendor selection?
• What is the failure cost per hour?

This is the step most people skip, and it is the one that causes the most harm.

Step 2: Suspect Your Sulzer Electronics Before You Tear Down the Pump

Most "emergency" mechanical failures I have been called to are actually instrumentation faults. Sulzer electronics—protective relays, vibration monitors, temperature transmitters, governor controls—can mislead you if they have drifted.

This brings us to the question that shows up in search logs: "What is the theory of drift?" In control systems, drift is the slow, unintended change in a signal over time. It is not a sudden spike. It is a gradual shift—like a pressure transmitter that outputs 4.3 mA when it should be 4.0 mA at zero pressure. Heat, vibration, and age cause it. It becomes a real problem when drift crosses an alarm setpoint and shuts down a perfectly healthy machine.

So before you order a replacement pump, ask your technician to:

1. Record all live readings and compare them to the last baseline.
2. Check sensor supply voltage and loop calibration.
3. Look at trend data—did the signal shift slowly or jump instantly?
4. Test the sensor against a known reference, not just "looks okay."

A 30-minute calibration check has saved me three full teardowns. The most frustrating part of this job: the same sensor drift issues recur despite clear documentation. You would think written specs would prevent it, but technicians interpret them differently.

Step 3: Gather the Exact Machine Identity

For a Wärtsilä-Sulzer RTA96-C dizel motor—or any large Sulzer rotating machine—"I need a seal" is not a sufficient request. You need:

• Full model and serial number, stamped on the nameplate, not in the manual.
• Pump or engine arrangement number, if available.
• Component revision or part number from the parts catalog.
• Bore size, shaft diameter, material code, and any special build notes.

I once ordered what I thought was the "standard" split ring for a Sulzer pump. The supplier also thought standard. We both meant different things. Discovered this when the part arrived and the fit was wrong. That was a $3,500 mistake and a 36-hour delay. Note to self: always photograph the nameplate and email it before you confirm the order.

Step 4: Choose a Sourcing Strategy, Not a Vendor Hope

For emergency parts, you need two options in parallel. If you only call one supplier, you are not managing risk—you are praying.

Our approach:

• OEM Sulzer service network for genuine parts and technical support.
• An independent rotating equipment shop for machining or fabrication.
• A third "wildcard" only if the first two can't meet the deadline.

And do not treat small orders differently. I worked with a startup that needed a single $300 bearing housing for a Sulzer pump. The first supplier said, "We will get to it when we can." The second treated the order with respect. Guess which one got the client's $18,000 overhaul three months later? Small doesn't mean unimportant—it means potential.

Part of me wants to consolidate to one source for everything. Another part knows that redundancy saved us in an emergency. I compromise with a primary-plus-backup approach.

Step 5: Make the Decision with a Time-Stamped Log

When the clock is ticking, everything becomes verbal. That is how mistakes happen. Start a simple log:

• When did you contact each vendor?
• What delivery promise did they make, and in what time zone?
• What is the worst case if you wait for this option?
• Who made the final call, and at what time?

I had two hours to decide between an OEM part with a known lead time and an aftermarket part that might arrive sooner but with lower confidence. Normally I would get three quotes and run a comparison. No time. Went with the vendor who answered the phone and provided a serial-number-matched quote. In hindsight, I should have asked for the calibration certificate before agreeing. But with the plant down and a CEO waiting, I made the call with incomplete information. The log documented why, so nobody had to learn the lesson twice.

Step 6: Add a Buffer and a Backup Plan

The vendor says: "It will be there by Friday." You should hear: "It will probably be there by Friday, but there is a chance it will not." Recalculate your schedule with a 20-30% buffer. If the promised lead time is five days, plan for six and a half. Buffer time is not ideal. It is necessary.

The third time a rush order arrived late from a normally reliable supplier, I finally created a formal buffer policy. Should have done it after the first time.

A decision without a backup is a gamble. A backup without a trigger time is paperwork.

Step 7: Verify Before You Install

Fast doesn't mean careless. When the part lands, do a quick inspection:

• Compare part numbers against the purchase order, not the packing slip.
• Check dimensions, bore size, and shaft fit with a caliper or micrometer.
• If it is a pressure part, ask for the material certificate per EN 10204 3.1.
• Test-run the electronics on the bench or with a signal simulator.
• Check for shipping damage—a bent shaft can hide inside a perfect box.

One more thing: if you searched for "eddie outlet" and found this page, you probably meant an eddy-current probe. The bearing outlet is where an eddy-current probe reads shaft position. Search the correct term and you will find the correct spare part.

If this is a hydrocarbon pump, API 610 is the standard to reference. It expects a minimum service life of 20 years and at least three years of uninterrupted operation. If you are replacing a component every year, the emergency is a symptom, not the problem.

After the Fix: Three Things That Go Wrong Next

1. You forget to update the spare part min/max levels. The emergency part that saved you should be in your stockroom now, not reordered in a panic next time.
2. You don't calibrate the rest of the loop. Drift doesn't stop because you replaced one sensor. Check adjacent sensors for baseline shift.
3. You close the work order without documenting what actually happened. The next person facing a similar Sulzer pump failure deserves your handwritten notes—especially the wrong assumptions.

The last one is personal. Our company lost a $40,000 maintenance contract in 2023 because we tried to save $600 on a standard inspection instead of a rush inspection. The pump failed again, and the client did not want to hear about our budget logic. That is when we implemented our photo + serial number + written acceptance policy. Simple. Expensive lesson.