What a Hydraulic System Is Actually Trading Away

Why this matters

Hydraulics buys you force in a small package and stiff, controllable motion, and it sends four bills for it: heat, dirt sensitivity, energy that survives the off switch, and pressure that goes through skin. Three of those four are invisible until something has already gone wrong. The first one is the one that decides whether a machine runs all shift or shuts down at two in the afternoon in July, and it is fully calculable at the machine with a pressure gauge and a flow figure off the pump nameplate. This card is that arithmetic and what the other three cost.

What the pressure buys you

Force out of a cylinder is pressure times piston area, and pressure is the cheap variable. A cylinder with a 3 inch bore has a piston area of 7.07 square inches. At 2,000 psi it delivers 14,140 lbf out of a tube a little over 3 inches across.

Get the same force from compressed air at 100 psi and you need 141 square inches, which is a 13.4 inch bore. Twenty times the area, four and a half times the diameter, and a cylinder that will not fit where the hydraulic one fits. That ratio is the entire commercial case for fluid power, and it holds anywhere the working pressure ratio holds.

You also buy two things that are harder to see. Stiffness: oil is nearly incompressible, so a hydraulic actuator holds position under a changing load in a way a pneumatic one cannot. And transmission: power routes through hose around corners that a shaft or a belt cannot turn.

Bill one: the loss does not go anywhere, it becomes heat in the oil

This is the claim the rest of the card rests on. Hydraulic power in gallons per minute and psi converts as flow times pressure divided by 1,714 to horsepower. Any of that power which does not end up moving a load ends up as heat, deposited in the fluid at the exact restriction that caused it: across a relief valve, across a throttling orifice, across a partly shifted spool, along an undersized hose.

There is nowhere else for it to go. Unlike an electric motor, which sheds loss to the air off a large frame, a hydraulic circuit puts its entire loss into a few gallons of liquid that then has to carry it back to a reservoir and give it up there. So reservoir temperature is a direct read of circuit efficiency, and an oil temperature that climbs through a shift is not a cooling problem first, it is a circuit that is doing something other than work with its power.

The reservoir is a poor heat sink and the numbers are not close. Petroleum hydraulic oil runs about 7.2 lb per gallon with a specific heat around 0.45 BTU per pound per degree F at operating temperature, so a 20 gallon charge is 144 lb and 64.8 BTU per degree F of thermal capacity. One horsepower is 2,545 BTU per hour. A single horsepower of continuous loss into that reservoir, with no heat rejected at all, raises the oil about 39 F per hour. Those fluid properties are for petroleum oil; a water-glycol fire-resistant fluid has a considerably higher specific heat and a different density, so the same heat input produces a slower rise and that fluid's own datasheet governs the numbers.

An uninsulated bare steel reservoir in still air sheds heat at roughly 1 to 1.5 BTU per hour per square foot per degree F of oil-to-air difference. That rule of thumb is derived for bare steel with no fan on it and no insulation; a painted, jacketed or shrouded tank sheds less, and the machine builder's own thermal sizing owns the real figure. It is a small number either way, which is why machines that lose more than a fraction of a horsepower continuously carry a cooler.

Bill two: the clearances are the same size as the dirt

The working clearances inside a piston pump, a servo valve or a proportional valve are measured in single-digit micrometres, which is the same size range as particles you cannot see, cannot filter out with a coarse strainer and will not notice on a dipstick. That is why contamination is the dominant failure cause in this discipline by a wide margin over every mechanism a tech is more likely to suspect. The sibling article on contamination owns the mechanism, the cleanliness code arithmetic and the sampling method; do not re-derive it. What belongs on this ledger is only that the sensitivity is a consequence of the same choice that bought you the force density: high pressure across small clearances is what makes a compact pump possible, and small clearances is what makes dirt lethal.

Bill three: the energy outlives the off switch

An accumulator, and to a lesser degree any long run of hose and any trapped cylinder volume, holds usable energy after every power source is isolated. Killing the motor does not de-energize it. That makes stored energy a separate isolation step rather than a consequence of the first one, which is exactly how 29 CFR 1910.147 treats it: at (d)(5)(i) the standard requires that after the energy isolating devices are locked out, all potentially hazardous stored or residual energy be relieved, disconnected, restrained or otherwise rendered safe. The companion articles on what an accumulator stores and on discharging one before work begins carry the arithmetic and the procedure.

Bill four: the pressure goes through skin

Fluid escaping a pinhole in a hose or a fitting is a needle you cannot see. Skin penetration is commonly cited from around 100 psi upward, which is a small fraction of the working pressure of every circuit discussed here, and no glove made resists it.

Never search for a hydraulic leak with a hand, and never run a hand along a hose to find a spray. The correct method is to shut the machine down and relieve pressure first; where a leak genuinely must be located under pressure, it is located with a piece of stiff cardboard held at arm's length, with the body out of line and eye protection on, and never with skin.

A fluid injection injury presents as a trivial puncture with almost no bleeding and often very little pain in the first hour. It is a surgical emergency, it is routinely under-triaged because it looks like nothing, and delay costs tissue and sometimes the digit. Anyone with a suspected injection goes to an emergency department immediately, says the words "high pressure fluid injection injury" rather than describing a puncture, and takes the safety data sheet for the fluid with them.

Worked example: the energy balance on a clamp cycle

One machine, one gauge, one nameplate. A fixed-displacement pump delivering 8 gpm, a relief valve set at 2,000 psi, a 20 gallon reservoir with roughly 22 square feet of wetted surface, and a 3 inch bore cylinder that clamps a part.

The cycle: extend for 5.5 seconds against a load that needs 800 psi at the cylinder, with 100 psi of line and valve loss, so the pump sees 900 psi. Then hold the clamp for 40 seconds with the directional valve centered and all four ports blocked, so the full 8 gpm goes over the relief valve at its setting. Then retract for 4.1 seconds. Call it a 50 second cycle, 72 cycles per hour.

Extend, with the useful work subtracted before the heat is counted. Total: 8 gpm times 900 psi divided by 1,714 is 4.20 hp. The part doing work on the load is 8 times 800 divided by 1,714, or 3.73 hp. The heat is the difference, 0.47 hp, for 5.5 seconds. At 0.7069 BTU per second per horsepower that is 1.83 BTU.

Hold. Nothing moves, so no work is done and every bit of it is heat: 8 gpm times 2,000 psi divided by 1,714 is 9.33 hp, for 40 seconds, which is 264 BTU.

Per cycle: 266 BTU, of which the 40 second hold contributes 99 percent. The clamp is not the load on this machine. The hold is.

Per hour: 266 times 72 cycles is about 19,150 BTU.

The rise, stated as the bound it is. Against 64.8 BTU per degree F of oil, with no heat rejected at all, the oil temperature rise is < 295 F per hour, which is under 5 F per minute. That is an upper bound and only an upper bound, because rejection has been left out deliberately and because the rate falls as the oil warms and the tank sheds more.

Now add the rejection term back. At the midpoint of the rule of thumb, 1.25 BTU per hour per square foot per degree F over 22 square feet, the tank sheds 27.5 BTU per hour for every degree F the oil sits above ambient. To shed 19,150 BTU per hour the oil would have to run 696 F above the room, which no fluid and no seal survives. The tank cannot get there, so the machine climbs until something trips or something fails.

The same machine with the hold term removed. Center the directional valve on a condition that routes pump flow back to tank at low pressure during the hold instead of blocking it, or drive the circuit with a pressure-compensated pump that destrokes when nothing is moving. The hold contributes almost nothing, leaving 1.83 BTU per cycle, or about 132 BTU per hour. Divided by the same 27.5 BTU per hour per degree F, the oil settles about 4.8 F above ambient. The bare tank handles it with no cooler at all.

Same pump, same cylinder, same duty, same 8 gpm. One choice about what happens during the 40 seconds when nothing is moving is the difference between a machine that never gets warm and a machine that cannot be cooled. That choice is a center condition or a pump type, and both are covered in their own cards.

The failure mode. The field version of this mistake is a cooler quotation. The oil is hot, so a cooler gets added, sized against the measured temperature. It works for a season, then the cooler fouls slightly and the machine is hot again, and the second cooler is bigger. Nobody ever asked what was generating 19,150 BTU per hour on a machine whose real work is under four horsepower for eleven percent of the cycle. Measure the heat before you size the sink.

What flips the trade

• A duty that is mostly holding, not moving. The example is the general case: hydraulics is expensive to leave switched on. If the load is held far more than it is moved, either the circuit unloads during the hold or hydraulics is the wrong choice.
• A cold environment. Everything above assumes the problem is shedding heat. Below freezing, viscosity is the constraint instead, pressure drops rise, and the same losses that were a liability become the warm-up.
• A fire-resistant fluid requirement. Where a circuit runs near an ignition source, the fluid choice is made on flammability first, and its heat capacity, its compatibility with seal materials and its cost of contamination all change together. Route that to the machine builder and the fluid supplier.
• Force that could be produced another way. The 20 to 1 area advantage evaporates if the load is modest. An electromechanical actuator has none of these four bills, and the honest answer on a low-force application is often that hydraulics is being used because the machine already had a power unit.

References

• 29 CFR 1910.147, the control of hazardous energy, including the requirement at (d)(5)(i) to relieve stored and residual energy after isolation devices are locked out
• Safety data sheet for the specific hydraulic fluid in the machine, which governs its properties, its skin and eye handling, and what an emergency department needs to know after an injection injury
• Machine builder documentation for reservoir sizing, cooler capacity and permitted operating temperature, which own those values
• See related: Why Contamination Is the Dominant Failure Cause in Hydraulics; What an Accumulator Stores and Why It Survives a Lockout; What Pressure and Flow Each Decide in a Hydraulic Circuit