A sprinkler drawing that only shows heads on a ceiling plan is not a design. The design is the proof that water will come out of the worst heads at the rate the hazard needs, at the pressure the site actually has. That proof is the hydraulic calculation. It sits in the middle of installation and design work: after the occupancy and AS 2118.1 hazard are agreed, before anyone orders pipe, and again whenever a fitout moves a range.
Owners do not need to run the software. They do need to know what the pack is claiming, because a calculation that assumes a generous town main, or the wrong density, will look tidy and still fail on the night.
What is a hydraulic calculation?
In plain terms, the calculation picks a remote area — the group of sprinklers that is hardest to supply — and shows that those heads can throw the design density across that area. Density is millimetres per minute, the same number as litres per minute per square metre. Light-hazard office space wants a modest density over a modest area. Ordinary-hazard shops want more. High-piled storage and process risks want more again, sometimes with a different head altogether.
The remote area is the patch of ceiling where friction, elevation and the number of open heads demand the most from the supply. In a rectangular warehouse it is often the far corner. With a long dead-end range, a high-level plant deck or a pressure-reducing station, it may not be the corner you first circled. A competent pack shows the remote area on a drawing, lists every head in it, and carries pressure back through ranges, mains and risers to a single demand point: flow and residual pressure at the town-main connection, the pump discharge or the tank outlet.
Everything downstream of that point is plumbing. Everything upstream is a water-supply decision: can the street deliver that residual at that flow, or do you need a tank and a pump? The calculation does not invent water. It states how much the system will ask for.
Hand calculations still matter. Before listed software, designers walked the remote area with Hazen–Williams tables, equivalent lengths and a pad of paper. The arithmetic has not changed. Software is faster, but it will compute a fantasy if the hazard, K-factor or supply curve is wrong. A reviewer who cannot reconstruct the remote-area demand on a whiteboard cannot defend the stamp.
Remote area, design density and K-factor
Three numbers decide whether the system is in the right conversation.
The design density comes from the hazard. AS 2118 and the storage rules beside it do not let a designer invent a comfortable millimetre figure because the architect wanted smaller pipe. Change the commodity, storage height or ceiling height, and density and design area both move. A pack that still shows last year’s shop density after the tenant installed racking is not a calculation of the building you now have.
The remote area is the design area placed in the hydraulically worst location. Software will nominate one if you let it. You should still look at the plan. A remote area that ignores a mezzanine, a high-level void, or a range value-engineered to 25 mm next to 32 mm is how rejections start. In high-rise work there may be a remote area in each pressure zone, because a zone valve and a pressure-reducing set change the supply the heads can see.
K-factor is the head’s discharge coefficient. In the units Australian software uses, flow in litres per minute is K times the square root of pressure at the orifice. A standard 15 mm office sprinkler is commonly a K80. Larger orifices — K115, K160, K200 and ESFR sizes — throw more water at the same pressure, or the same water at a lower pressure. You cannot swap a head on site because the ceiling tile wanted a different finish. Change K and you change every head in the remote area, and you may change whether the town main is still enough.
The calculation also carries hose-stream or inside-hose allowances where the adopted method requires them. Those extra litres are for the fire service on the landing, not for the sprinklers. Leaving them out makes the graph look healthier than the street will be on the night.
Friction loss, velocity and the residual pressure graph
Water moving through steel, copper or listed plastic loses pressure to friction and fittings. Designers still use Hazen–Williams for the pipe, and equivalent lengths for grooved couplings, tees, valves and backflow preventers. Elevation is kilopascals per metre of rise. What is easy to hide is an optimistic C-factor (pipe roughness), a forgotten strainer, or a backflow-prevention assembly the authority will insist on and the model never included.
Velocity is the quiet limit. High velocity raises friction, noise and water-hammer risk when a valve slams. Listed fittings and the adopted standard keep velocities inside a band — in steel mains, designers commonly stay well below about six metres a second. If the only way to hit density is to run the remote range at a sprint, the pipe is too small or the supply is too weak. Opening the velocity limit in the software to “make it work” is a rejection waiting for a reviewer who reads that column.
The town-main residual pressure graph is the other half of the proof. A static pressure on a quiet Sunday morning is not a supply. The authority, or a contractor with a calibrated test, flows the main and records residual pressure at those flows. Plotted, that is a falling curve. The sprinkler demand is a point on the same axes. If the demand sits comfortably under the residual curve, with a margin for ageing mains, peak-hour demand and the next development on the street, a town-main supply can be argued. If it sits on or above the curve, you are buying a tank and a pump.
I have seen packs that quote a static figure from a neighbour’s building and never show a residual, and residuals taken at 9 am in a suburb that loses twenty metres of head at 7 pm. The graph is the only picture of the street the calculation can be compared with. High-rise zones and pumped supplies still need it, because the pump duty is chosen against the same conversation.
Software versus hand calculations
Listed hydraulic packages are now the ordinary tool on Australian jobs. They take a node model, apply K-factors and elevations, and iterate until every open head in the remote area meets density. Used well, they catch the range you would have undersized by hand. Used poorly, they launder a bad hazard into a colour plot.
Hand calcs remain the check. A reviewer should take the remote-area heads, assume a reasonable end-head pressure from K-factor and density, walk the friction back, and land near the software’s demand. If the two answers disagree widely, someone has a wrong equivalent length, a missing elevation, or the wrong remote area. For a small residential or FPAA system the whole proof may still be a short hand sheet. That is proportionate, not old-fashioned.
Software will not choose the system type. Wet, dry, pre-action and deluge change the demand: dry pipe needs extra pressure to push water through air-filled pipe; deluge assumes every head is open. If the model is still a wet-pipe ordinary-hazard file, the plot describes a building you are not building.
Why calculations get rejected — and what owners should ask to see
Rejections are rarely about a single mis-added metre. They cluster.
The hazard or density does not match the use. The remote area is not shown, or is not the worst patch. The town-main residual is a static number, a phone call, or a test older than the last street main relay. Hose-stream allowance is missing. Velocities are over the adopted limit. A pressure-reducing valve, a backflow preventer or a long authority service has no equivalent length. Heads on the drawing are a different K-factor from the heads in the table. A shared hydrant and sprinkler supply has been calculated as if the hydrants were asleep. Dry or pre-action delay has been ignored. The pack has no name, no date, and no revision after the fitout moved the ranges.
If you own the building, you do not need the node table. Ask for four sheets and read them in order:
- The hazard statement — occupancy, commodity, storage height, design density and design area.
- The remote-area plan, with the flowing heads marked.
- The demand summary — flow and pressure at the base of the system, plus hose-stream allowance.
- The water-supply graph, with the demand point plotted on a residual curve from a dated test, or on a pump curve that names the duty.
If those four sheets cannot be produced, there is no calculation to certify. A stamp on a ceiling plan is not a substitute. And if the fitout, the racking or the street main changes, the pack is a historical document until someone runs it again.