Lubrication System Design: Grease vs Oil, Delivery Methods, and Maintenance Intervals

Lubrication is one of those design decisions that gets made in five minutes during a component selection meeting and then determines how much maintenance headache a machine generates for the next fifteen years. I have worked as a contract mechanical designer inside large manufacturing plants in Japan for two decades now, and I have seen more unplanned downtime traced back to a poorly chosen lubrication strategy than to almost any other single design decision. The grease-versus-oil question, the delivery method, and the maintenance interval are not independent choices — they interact with each other and with the operating environment in ways that are easy to get wrong if you treat lubrication as an afterthought bolted onto a finished bearing or gear selection.

Choosing Between Grease and Oil for the Application

Grease is oil combined with a thickener, and that thickener is the whole story of why grease behaves so differently from straight oil in service. Grease stays put, seals out contamination reasonably well, and needs no reservoir or pump, which makes it the default choice for bearings that are hard to reach or that run in dirty environments. On a conveyor idler roller design, grease-packed sealed bearings meant zero ongoing lubrication maintenance for the life of the bearing, which was the right call given that the rollers were spread across a hundred-meter line and individually servicing each one with an oil system would have been impractical.

Oil, by contrast, carries heat away from the bearing far more effectively than grease because it can be circulated, filtered, and cooled, and it flows into tight clearances more readily. On a high-speed spindle project running well above the speed range where grease starts to churn and overheat, we moved to an oil-air minimum quantity lubrication system specifically because grease at that speed would have generated enough internal churning heat to cook itself and fail early. The speed factor, commonly expressed as the bearing bore diameter in millimeters multiplied by the shaft speed in rpm, is the number I check first when deciding between grease and oil — as a rough rule of thumb, once that DN value climbs into the higher ranges typical of machine tool spindles, grease’s practical speed limit starts to become the binding constraint, and oil lubrication becomes the more reliable choice regardless of how convenient grease would otherwise be.

I made the wrong call once on a gearbox project early in my career, specifying grease lubrication on a right-angle gearbox because it simplified the housing design and avoided a seal and reservoir, without checking the manufacturer’s speed and load rating tables closely enough. The gearbox ran hot within the first month of continuous operation, and grease that had been rated for the load alone could not handle the combination of load and the churning at that particular gear mesh speed. Switching to an oil bath with a sight glass solved the overheating, but it cost a redesign of the housing to add a proper reservoir and fill and drain plugs, work that would have taken a fraction of the time if I had done the speed and load check correctly before committing to grease in the first place.

Delivery Method: Matching the System to Accessibility and Duty Cycle

Once grease or oil is chosen, how it actually gets to the friction surface is its own design problem. For grease, the simplest method is a hand-packed bearing sealed for life, appropriate when access is difficult or the duty cycle is light enough that the grease’s working life exceeds the component’s service life. The next step up is a grease fitting with a manual grease gun on a scheduled interval, which works well when a small number of points need periodic attention and someone is already walking the machine regularly for other inspections. Beyond that sits automatic lubrication, either single-line resistance systems or progressive divider systems that meter grease to multiple points from a central reservoir on a timer.

On a stamping press retrofit, we had eleven grease points scattered across a machine that ran three shifts, and manual greasing on that schedule was simply not happening reliably — maintenance logs showed points going two or three times longer than the specified interval because operators were busy and the points were physically awkward to reach. Installing a progressive divider automatic lubrication system with a timer-driven pump solved the reliability problem completely, because it removed the human compliance step from the equation entirely. The upfront cost was higher than a fitting and a grease gun, but it was justified purely on the basis of the two bearing failures per year we had been experiencing from under-greased points, each of which cost far more in downtime than the automatic system’s installed price.

For oil systems, the delivery method spans a similarly wide range, from a simple splash system where rotating gears sling oil around an enclosed housing, to a forced circulation system with a pump, filter, and cooler for high-load or high-speed applications. Splash lubrication is elegantly simple and needs no moving parts beyond the gears themselves, but it depends entirely on maintaining the correct oil level and on the gears actually reaching into the oil at the right depth, which is a detail that is easy to get wrong in a housing redesign that changes gear position without rechecking the oil level line. I have seen a gearbox redesigned for a slightly different ratio where the lowest gear no longer dipped into the oil sump at the new center distance, and nobody caught it until the gearbox failed from inadequate lubrication a few months into service.

Delivery Method Comparison

Setting Maintenance Intervals That Match Real Operating Conditions

Manufacturer-published relubrication intervals assume a baseline operating environment, and applying them blindly to a harsher or cleaner-than-assumed environment produces either premature failures or wasted maintenance effort. Temperature is the biggest multiplier — grease life roughly halves for every rise of about fifteen degrees Celsius above the baseline temperature the interval was rated at, which means a bearing running hot next to a furnace or a motor housing needs a dramatically shorter interval than the same bearing in a climate-controlled area, even though the catalog number is identical.

On a furnace-adjacent conveyor drive, I initially specified the standard catalog relubrication interval for the bearing, and it failed well before that interval was up. Pulling the failed bearing showed classic signs of thermally degraded grease — hardened and darkened rather than the smooth consistency of fresh grease. Recalculating the interval using the manufacturer’s temperature correction factor for the actual measured housing temperature, which was running considerably hotter than the catalog baseline assumed, cut the recommended interval substantially, and switching to a higher-temperature-rated grease alongside the shorter interval eliminated the repeat failures. The lesson that stuck with me is to always measure or estimate actual operating temperature at the bearing, not the ambient shop temperature, because the two can differ enormously near heat sources.

Contamination exposure cuts the interval just as sharply as heat does, and it is easy to underestimate for equipment running in a dusty or wet environment. On outdoor equipment exposed to washdown, water ingress past aging seals contaminated the grease well before the calendar-based interval would have called for a change, and the fix was not just a shorter interval but also a seal upgrade to a contact lip seal rated for washdown, paired with more frequent grease sampling during the first year to validate that the new seal was actually keeping water out before relaxing the interval back toward standard.

Contamination Control and Filtration as Part of the Lubrication Design

For oil systems especially, filtration is not a maintenance add-on, it is a design decision that determines component life directly, because particle contamination is one of the leading causes of premature bearing and gear wear in circulated oil systems. On a hydraulic power unit, undersizing the return line filter to save cost seemed like a minor decision at the time, but it allowed particulate from initial system break-in to circulate through the pump and valves for weeks before anyone noticed degraded performance, and the resulting wear on the pump’s internal components was expensive to repair compared to what a properly sized filter would have cost upfront.

I now specify filtration based on the most sensitive component in the circuit, not an average across the system, because a single tight-clearance servo valve downstream of a coarse filter will fail from contamination that a less sensitive gear pump would never notice. Breather design matters just as much and gets ignored even more often — an open breather on an oil reservoir lets moisture and airborne dust into the system with every thermal cycle as the reservoir breathes in and out, and switching to a desiccant breather on a reservoir that had been experiencing water contamination in oil samples solved a problem that had been misdiagnosed as a seal leak for months before someone thought to check the breather.

Building Lubrication Into the Design Review, Not the Punch List

The pattern across every lubrication failure I have investigated is the same: the lubrication strategy was decided quickly, early, and in isolation from the rest of the design, then never revisited as the design matured around it. I now treat lubrication method and interval as a line item in every design review, right alongside load capacity and material selection, and I ask explicitly whether the chosen method matches the actual accessibility, duty cycle, temperature, and contamination environment the machine will see in the field, not the idealized environment assumed in a catalog table. That single habit, forcing the question into the review rather than letting it default silently to whatever the bearing supplier’s standard recommendation happened to be, has prevented more repeat failures on my recent projects than any other single change I have made to my design process.