Episode 75: Why Service Intervals Matter More Than You Think

Jason Reed

[excited] Welcome to the show -- and I gotta say, Lisa, this one kind of sneaked up on me. Episode 75. Seventy-five! [laughs] The diamond anniversary, apparently. Which feels way too fancy for two people talking about compressors and maintenance, but here we are.

Lisa Saunders

[warmly] Yeah, 75 is the part that gets me. Not five, not 10, not even 50 -- SEVENTY-five. That's enough episodes to make me think, okay, somebody out there actually wants to hear practical shop talk. Which, honestly, I love. Because the stuff that keeps plants running is usually not dramatic. It's routine. It's checklists. It's somebody remembering to look at the thing nobody feels like looking at.

Jason Reed

Exactly. Nobody throws a party because a condensate drain got cleaned on schedule. Nobody hangs a banner because somebody changed a filter at 4,000 hours. But miss that boring little job, and now everybody's standing around a dead compressor asking how this happened.

Lisa Saunders

[questioning tone] And that's the trap, right? We talk about reliability like it's this big strategy deck, but half the time it's one ugly truth: did you do the maintenance when you said you were gonna do it? Or did the reminder sit there until it became a problem with a price tag?

Jason Reed

[matter-of-fact] Right. Service intervals sound like paperwork. They're not paperwork. They're the line between planned maintenance and expensive surprises. If you've got an air filter due, oil coming up on hours, drains not being checked, leaks ignored -- you're not saving time. You're borrowing trouble.

Lisa Saunders

And usually at the worst possible moment. [sighs] Not Tuesday at 10 a.m. when everybody's staffed up. It's Friday night, or during a production run, or right when the plant can't afford to lose air. Compressed air has this nasty habit of becoming visible only when it's gone.

Jason Reed

That's a good way to put it. When the air's there, nobody says a word. When it's not, suddenly compressed air is the most important utility in the building. So if you're responsible for that system, the calendar matters. The hour meter matters. The service log matters.

Lisa Saunders

[curious] But I do think people hear "service interval" and imagine a calendar on the wall -- first Monday, change this; third Thursday, check that. And some of it is calendar-driven, sure, but operating hours matter too. A compressor working hard tells the truth faster than the date does.

Jason Reed

Yep. Track both. Hours tell you how hard the machine is actually running. Calendar time matters because oil ages, seals age, moisture shows up even when the machine isn't living a heroic life. So this isn't just, "Did we remember a date?" It's, "Are we paying attention before the machine teaches us a lesson?"

Lisa Saunders

[lightly] Which is a very expensive teacher.

Jason Reed

[chuckles] Tuition is brutal. And that's really the whole setup for today: service intervals are necessary, but if you treat them like a box to check and nothing more, you're leaving a lot on the table -- and taking on risk you don't need.

Lisa Saunders

[curious] Okay, so let's get into the part people can use. When you say service intervals are necessary but not enough, what are the absolute must-dos?

Jason Reed

[matter-of-fact] Start simple. Change air filters every 4,000 hours, or sooner if the indicator tells you they're loaded up. Check oil level before startup. Clean condensate traps and drains daily. Use the oil and lubricants the OEM calls for -- not whatever was sitting on a shelf because it was cheap. Change the oil separator and lubricant every year or every 8,000 hours, depending on the machine guidelines. And if you're using food-grade lubricant, tighten that up -- more like every 4,000 hours.

Lisa Saunders

Wait -- that 4,000-hour split for food-grade lubricant is exactly the kind of detail people miss. They hear "oil is oil," and it is absolutely not.

Jason Reed

Correct. And that's where folks get themselves in trouble. The manual gives you a baseline, but your plant conditions may not let you run that far cleanly. Dust in the air, dirt from outside, caustic chemicals, high heat, moisture, short cycling -- all of that changes the conversation.

Lisa Saunders

[skeptical] So let me push on that. If the manual says 8,000 hours for oil, and a plant manager says, "Great, see you at 8,000," you're saying that's a minimum rule, not a promise.

Jason Reed

That's EXACTLY what I'm saying. OEM intervals are minimum recommendations to keep the machine operating as designed and to protect warranty eligibility. They are not a magical guarantee that every facility gets the same result. A clean climate-controlled room and a dusty plant next to a racetrack are not living the same life.

Lisa Saunders

[laughs] The racetrack example sticks, because you can picture it. Fine dust everywhere, getting pulled into the inlet air all day. Of course that machine is gonna want more attention than the one sitting in a cleaner environment.

Jason Reed

Exactly. And then there are leaks. People treat leaks like background noise. They're not. Repair system air leaks as quickly as possible. You're paying to compress that air. Letting it escape is just burning energy for no reason.

Lisa Saunders

And it's sneaky because a leak doesn't always shut you down dramatically. It just drags on efficiency, drags on runtime, drags on cost. It's death by a thousand hisses.

Jason Reed

[deadpan] That might be the most accurate sentence in this episode. Now, the other piece is predictive maintenance. Preventive maintenance says, "At 8,000 hours, change the oil." Predictive maintenance says, "Let's look at what the machine is telling us before something goes sideways."

Lisa Saunders

So this is where oil sampling comes in.

Jason Reed

Yes. Oil sampling is one of the cleanest examples. Pull a sample every 2,000 hours -- and even more often in harsher environments or with food-grade lubricant, more like every 1,000 hours. That report can show elevated wear metals from bearings, rotors, gears, bushings. It can show viscosity issues, additive depletion, abnormal acid number, contamination, even water in the oil.

Lisa Saunders

[sharper implication] And water in the oil is not just a lab curiosity. If you see water, now you're asking about moisture, short cycling, maybe internal leaks -- real operating problems, not abstract chemistry.

Jason Reed

Bingo. Same with contamination. If a sample comes back dirty, that can point to poor filtration or lousy inlet air quality. That's predictive maintenance doing its job: catching a problem while it's still information instead of a failure.

Lisa Saunders

But -- and here's the tension -- people hear that and immediately go, "Great, if the oil sample looks good at 8,000 hours, I can skip the scheduled service." That's the part I don't buy.

Jason Reed

[firm] You shouldn't buy it. Predictive maintenance does NOT replace preventive maintenance. If the manufacturer says change the oil at 8,000 hours for warranty and reliability, do it. Even if the report looks decent. Because "decent right now" is not the same as "safe to ignore the rulebook."

Lisa Saunders

And warranty is the part people remember too late. Most manufacturers want proof you did the required service at the required times. No records, no proof, now you're having a very unpleasant conversation.

Jason Reed

That's right. And in some industries it's bigger than warranty. Think food and beverage. Filter maintenance there is not optional housekeeping. Even a small amount of oil carryover can ruin product, contaminate piping, foul downstream equipment. Replacing filters on time is cheap compared to cleaning up that mess.

Lisa Saunders

[reflective] So the smart version of this is kind of a hybrid. Keep the required intervals. Then layer in real data -- oil samples, pressure levels, CFM demand, energy consumption, load hours, run hours -- so you're not flying blind between those intervals.

Jason Reed

Exactly. Build a baseline. Know what normal looks like for your system. Then when pressure drifts, energy use climbs, moisture shows up, or runtime changes, you've got context. Data becomes a flashlight. It shows you where to look.

Lisa Saunders

Not a hall pass.

Jason Reed

[chuckles] Not a hall pass. That's the line. Data helps you tighten maintenance based on reality. It doesn't let you pretend the minimums don't exist.

Lisa Saunders

And honestly, that's the grown-up version of maintenance. Not reactive, not superstitious, not just changing parts because the moon is full. You respect the rulebook, and then you pay attention to your actual plant.

Jason Reed

[calm] That's it. The machine gives you a schedule. The environment edits it. And your job is to notice.

Lisa Saunders

[warmly] That's a good place to leave it. Thanks for being here for episode 75 -- still a little surreal. We'll see you next time.