Hydraulic machines almost never fail the way people expect. They don't stop - they slow down. And so gradually that nobody notices, and by the time anyone does, the loss has been baked into the production plan because it seems normal.

Here’s an example. A mesh line was commissioned at a 22 second cycle. Three years later, it runs at 26 seconds. Nothing has broken. There are no alarms, no oil leaks on the floor, no heat complaints, and the machine does exactly what the operator asks it to do. The only evidence is that the shift doesn't finish on time anymore. The difference is subtle and slow - therefore, managers can’t notice the difference.

What four seconds actually costs

Run the numbers on that line at 1,200 sheets per shift:

That 4 seconds has pushed the shift past its own length. You now have two options, and most plants are quietly taking the first one without ever having decided to.

Option one: run the overtime. About 1.3 hours per shift. Across two shifts a day and 240 production days, that's roughly 640 hours a year. Put a three-person crew on it at a loaded overtime rate of $65 an hour, and you are at $125,000 a year to produce exactly the output you used to produce inside ordinary hours.

Option two: accept the shortfall. Output drops to about 1,108 sheets a shift. That's 92 sheets a shift you aren't making, or around 44,000 sheets a year of capacity that has silently left the building.

Neither of those appears anywhere as a hydraulics cost. The overtime sits in the labour line. The shortfall sits in the production plan as the line's capacity, because that is what the line does now.

Why nobody catches it

Three reasons, and they compound.

There's no baseline. Almost nobody records cycle time at commissioning, and if they do, it's in a file nobody can find three years later. Without the original number, drift is invisible.

It never trips anything. A machine that is 18% slower is still fully functional. Every condition the control system monitors is within limits. Pressure is right, temperature is right, and nothing is leaking. The only parameter that changed is one nobody is watching.

It gets blamed on people. This is the expensive one. When a line stops hitting its numbers, the first explanations reached for are the afternoon crew, the new operator, the material supplier, and a run of harder bar. Plants have restructured shifts and unfortunately replaced supervisors over a problem that was an accumulator losing nitrogen.

So what is actually happening?

Four mechanisms, all gradual, none of them visible from outside the machine.

1. Pump volumetric efficiency decline. As internal clearances open with wear, more of each revolution slips back inside the pump instead of reaching the cylinder. The pump still makes pressure—it just delivers less flow doing it. This is the most common cause and the most under-diagnosed.

2. Accumulator pre-charge loss. Nitrogen migrates out slowly over months. On any machine using an accumulator for the fast approach stroke, a soft pre-charge means the machine loses its punch and the pump has to supply what the accumulator used to deliver. Cycle time stretches immediately. Checking pre-charge takes minutes, and almost no plant does it on schedule.

3. Internal valve leakage. Spool wear and seat erosion let a small percentage of flow cross internally. Externally the valve looks perfect. There is nothing to see.

4. Cylinder seal bypass. Fluid crossing the piston inside the cylinder. It shows up as a slower, softer return stroke—not as oil on the floor, which is what people look for.

So, how do you tell hydraulic drift from everything else?

Fair question, and the test is simple. Hydraulic drift is consistent. It is the same on every shift, with every operator, on every grade of material, and it’s worse on unloaded fast strokes than under full load—because that's where flow, not pressure, is the limiting factor.

If the slow cycles cluster on one crew or one product, it isn't the hydraulics. If they're uniform across everything, it is, and no amount of supervision will fix it.

What to do about it, starting in November

Baseline every machine now while they're healthy. Cycle time and the time of each individual stroke, if your controls will give it to you. This costs nothing and is the single highest-value thing in this article. Without the baseline, you can never prove the drift exists, and you can never prove you fixed it.

Check accumulator pre-charge on a schedule. It's the cheapest of the four causes to find and the cheapest to correct.

Compare flow, not pressure. Pressure testing is what most people do, and it tells you very little about speed. A pump with badly degraded volumetric efficiency will still hit its pressure setting. Flow testing under load is what exposes it.

The reason this is worth doing is the ratio. Restoring that mesh line—pump, accumulator, valve overhaul—is typically a job in the twenty-thousand-dollar range. Set against $125,000 a year in overtime, the payback is measured in weeks, and you get the capacity back without buying a machine.

The capacity you lost is still in there. You just have to go and find it.

Final Thought

Flotek fixes hydraulic systems for Melbourne's steel processing, fabrication, and recycling sector. We can baseline and flow-test your critical machines and tell you where your cycle time has gone—before you spend money on capacity you already own.  And of course, this proactive improvement is found in all different industries where hydraulics is the key player.