Your central hydraulic power unit (HPU) will fail. The decision you make that week, will last twenty years.

Most steel plants in Melbourne are running hydraulics that were specified before the current maintenance manager started. The power unit often sits in its own room, or a fenced corner of the shed. Steel lines run out from it across the floor to the shear, the press, the straightener, the coil handling gear, etc. It may have been rebuilt once, maybe twice. The pump has been changed. The tank has likely never been drained properly, because there has never been a shutdown long enough.

And one day it stops.

What happens next is almost always the same. Production is down. Someone rings a hydraulic service company like Foltek. After diagnosing that repair is not viable, you request a like-for-like rebuild quote. And because time is ticking, you often make the decision to go ahead, because every hour of downtime is incredibly expensive. A twenty-year architectural decision gets made in four days.

That is the real problem. Not the failure - the timing of the decision.

What you are actually deciding

The question is not "hydraulic or electric." That framing is a sales pitch, and anyone who has watched a 400 tonne press hold load through a shift knows why it is wrong. Hydraulics still wins decisively on force density, on shock tolerance, and on holding a load in position without drawing power. In heavy production, those are not nice-to-haves. There is no electromechanical equivalent that survives a cropping shear's impact loading at the same cost.

The question is where the power gets generated.

The traditional answer is: in one place, at one pressure, distributed to everything. That architecture made sense when variable speed drives were expensive and controls were dumb. It has three weaknesses that have quietly got worse over thirty years.

The first is energy:

A fixed-displacement pump running a constant-pressure ring is producing full flow whenever it is switched on, and dumping most of it over relief for the large fraction of the shift when nothing is actually moving. That energy comes back as heat, which is why you have a cooler, and why the cooler is working harder every summer. With current industrial electricity pricing, that is no longer a rounding error on a plant's power bill.

The second is leaks:

Every metre of pressurised line is a leak path, and a plant with a central unit has hundreds of metres of it plus every fitting, and every one of those fittings is above a floor that eventually drains somewhere. Oil on the floor used to be housekeeping. Under EPA Victoria's general environmental duty, it has become a serious issue.

The third is concentration of risk:

One unit down means every machine it feeds is down. If your shear, your press and your straightener all run off the same ring, you do not have three machines - you have one machine with three heads.

The decentralised alternative

The modern answer is to keep hydraulic actuation and kill the central unit. That takes two forms, and they are not the same thing.

Variable-speed compact power units put a small motor-pump group at each machine, with the drive controller mounted on the unit itself. The motor only turns when the machine needs pressure, and it turns at the speed the duty requires. Manufacturers quote energy savings up to 70% depending on cycle - the real number depends entirely on your duty cycle, and for a machine that sits idle between plates it can be very large. The pipe run shrinks from a hundred metres to two.

Self-contained actuators go further and integrate the motor, pump and cylinder into one sealed assembly with no external lines at all. Loads are held by integrated shut-off valves with the motor stationary. These are genuinely impressive, and they are also the option that demands the most caution - they are typically limited to smaller power classes, they are not field-repairable in the way a conventional cylinder is, and if the manufacturer has no depth of support in Australia you have created a new single point of failure with a long lead time attached to it.

That last point matters more than any specification sheet. A sealed actuator with excellent European support and nobody in Melbourne who can look at it is a worse outcome than the pump you already know.

What to do about it now

The whole argument only works if the decision is made before the failure. Three things are worth doing in a quiet month:

Find out what the unit is actually costing you.

Metered power draw on the HPU circuit across a normal week, compared against the hours the machines were actually cutting. Most plants have never measured this and are surprised by it.

Get the drawings to reflect reality.

The schematic on file is almost certainly the original, and there have been a dozen undocumented modifications since. This is worth doing regardless of what you decide, and it is worth doing while the people who made those modifications still work here.

Decide the architecture in advance, then execute it in stages.

You do not have to convert the plant in one shutdown. The workable path is to decentralise machine by machine as each one comes up for major overhaul, with the central unit progressively carrying less load until decommissioning it is a small job rather than a capital project.

The plants that handle this well are not the ones with the newest equipment. They are the ones who decided what they were going to do about the power unit (HPU) two years before it stops working for good.

Flotek services and rebuilds industrial hydraulic systems across Melbourne's steel production and fabrication sector. If you would like an honest assessment of your central power unit, it’s remaining productive lifetime, and your options when it does finally stop working, get in touch.

Our Technicians have been trained in all aspects of HPU’s, so know them inside-out.