Nine litres a minute at the kitchen cold
The complaint arrives as a story rather than a number. The shower is fine at seven in the morning and useless at eight. Someone fills the kettle and the person upstairs feels it. A combi boiler cuts out mid-shower because its flow switch drops below the rate it needs to fire. None of that describes a pressure fault, although every householder in the country calls it one, and a good half of the trade prices it as one.
Two quantities are being confused, and separating them is most of the job. Pressure is what the water has when it is standing still, and it is set by the height of the service reservoir above the property. Flow is what the pipe will actually pass, and it is set by how much of that pressure gets eaten by friction on the way. A property can sit under a perfectly healthy static head and still deliver nine litres a minute at the kitchen cold, because the static reading is taken with nothing moving and the friction that ruins the job only exists once the water moves.
So the first visit is a measuring visit, and it produces three readings rather than one. Static pressure at the outside boundary stop tap with every outlet in the house shut. Then flow and residual pressure at that same point with a known draw running, caught in a graduated container against a watch. Then the same pair repeated at the internal stop tap and at the kitchen cold. The difference between the boundary figure and the internal figure is the supply pipe's contribution, and the difference between the internal figure and the tap is the house's. Until those three are on paper, any equipment discussion is a guess with a price on it. It is worth knowing that Ofwat's long-standing reference level for reporting a property at risk of low pressure is a static head of seven metres at the boundary at a flow of nine litres a minute — a threshold set for regulatory reporting, not a promise that a house meeting it will run two showers. A supply can be entirely within the undertaker's obligations and entirely unusable at the same time.
Where the pressure goes between the street and the tap
Write the budget down in the order the water meets it, because that order is also the order in which each item can be attacked. Start with the static head the undertaker will guarantee at the boundary, taken at the worst hour of the worst season rather than at the hour you happened to test. Subtract the lift to the highest outlet, at very close to 0.098 bar for every metre climbed. Subtract the loss through the meter and its check valve at the flow you actually want. Subtract friction in the underground service. Subtract friction and fittings in the internal distribution. What is left has to exceed the residual working pressure the worst appliance needs, and for a combi or a thermostatic mixer that residual is a manufacturer's figure, not a rule of thumb.
The item that surprises people is friction, because it does not scale the way intuition expects. In the Hazen-Williams form the head lost rises with flow raised to the power 1.852, so doubling the draw through a pipe multiplies the loss by roughly three and a half. That exponent is the entire explanation for the shower that dies when a tap opens. At nine litres a minute the old service is merely disappointing; at twenty it is a closed valve. Nothing has changed except the rate, and a static pressure test at the same tap five minutes earlier said everything was fine.
Run the arithmetic on a real service and the scale of it stops being an argument. Take a twenty-two metre run from the boundary to the internal stop tap — an ordinary front garden and a hallway — and ask for twenty litres a minute, which is roughly a shower and a basin tap together.
| Service pipe | Internal bore | C | Head lost | Equivalent pressure | Velocity |
|---|---|---|---|---|---|
| 15 mm lead, tuberculated | 12.7 mm | 100 | 29.0 m | 2.84 bar | 2.63 m/s |
| 15 mm copper, some scale | 13.6 mm | 130 | 12.8 m | 1.25 bar | 2.29 m/s |
| 20 mm MDPE | 16.0 mm | 150 | 4.4 m | 0.44 bar | 1.66 m/s |
| 25 mm MDPE | 20.4 mm | 150 | 1.4 m | 0.13 bar | 1.02 m/s |
| 32 mm MDPE | 26.0 mm | 150 | 0.4 m | 0.04 bar | 0.63 m/s |
Put the measured length, the real bore and an honest roughness in and the service pipe either explains the complaint or it does not — which decides whether the next purchase is a pump or a trench.
The total length of the pipe run.
The design flow rate through the pipe.
The pipe material's Hazen-Williams roughness coefficient.
The pipe's actual internal (bore) diameter.
Friction head loss
17.96 ft
They open the calculator with your figures already in it
Domestic Water Pipe Friction Loss Calculator (Hazen-Williams): 17.96 ft — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
What this calculation does not cover
- Hazen-Williams is an empirical fit to cold water in turbulent flow, roughly 40 to 75 °F (4 to 24 °C) at velocities in the 2 to 10 ft/s (0.6 to 3 m/s) band. Outside that it drifts: glycol-charged loops, hot recirculating mains and low-flow trickles are Darcy-Weisbach problems, where viscosity enters the calculation instead of being absorbed into a single C value. The formula still returns a number for any of them; it is just not the loss you will measure.
- Head loss is one term of the pressure the fixture actually sees. What is available is the incoming main, less the static lift to the outlet, less this friction, less the fittings, and less every device in the line — a meter, a backflow preventer, a filter and a softener can take 20 to 30 psi (207 kPa) between them. A run whose friction loss looks comfortable here can still open at a shower with nothing left.
- Velocity is not reported, and velocity is often what sizes the pipe rather than head loss. Copper is normally held near 8 ft/s cold and 5 ft/s or less on hot recirculating lines, because water moving faster erodes the inside of elbows and tees over the years and makes the pipe audible in the wall. A small pipe on a short run can pass a friction check comfortably and still be the wrong size.
Why the table settles most jobs on its own
Twenty-nine metres of head is roughly 2.8 bar, and no distribution main in a British street has that much to give away on friction alone before the house has been reached. That row is not a marginal supply, it is a supply that physically cannot deliver twenty litres a minute at any pressure the undertaker is going to provide. Fit a booster set behind it and the pump will pull the residual pressure down towards zero trying to find the flow, hit its low-pressure cut-out or start cavitating, and the customer will have paid for plant that made the noise worse and the shower no better.
Three rows down the same job is finished. Twenty-five millimetre MDPE takes the loss from 2.84 bar to 0.13 bar, and the house that could not run one shower now runs two on nothing but the main. The velocity column matters as much as the head column: the lead service is moving water at 2.63 m/s, which is above what anyone would design for in occupied premises and is a large part of why the supply is audible in the hallway. Sizing methods in BS EN 806-3 are built around velocity precisely because noise and erosion are governed by it independently of whether the pressure arithmetic closes.
Be honest about the roughness coefficient, because it is the one input on this calculation nobody can measure directly. A C of 150 belongs to new plastic. Old lead with a century of tuberculation inside it is not 150 and there is no published figure that is right for a specific pipe; 100 is a defensible pessimistic assumption and the sensible move is to run the calculation twice, once optimistic and once pessimistic, and see whether the conclusion changes. If both runs say the service is the problem, the diagnosis is safe. If they disagree, the answer is to dig a test hole and look at the pipe rather than to pick the number that suits the quote.
Two field checks catch the cases the arithmetic will not. Part-renewed services are common: someone laid 25 mm MDPE from the boundary to the wall in 2009 and left four metres of the old lead in the ground under the pavement, so the good pipe is in series with the bad one and the measurement at the internal stop tap looks inexplicable. And shared supplies still exist on Victorian terraces, where one service feeds several properties off a common spine and the flow available to any one of them depends on what the neighbours are doing. Both show up as a boundary reading that will not reconcile with the pipe you can see.
Renewing the pipe before pricing the plant
The underground run is in two halves with two owners, and quoting for the wrong half is a well-worn way to lose money. From the main to the boundary is the communication pipe and it belongs to the water undertaker. From the boundary into the building is the supply pipe and it belongs to the property owner. The boundary stop tap sits at the join and is the reference point for every measurement above. Most undertakers will renew or upsize their half on request and many operate lead replacement schemes that will do it free or at cost when the customer replaces theirs at the same time, which turns a service renewal into a considerably better job for a small amount of paperwork done in the right order.
For the customer's half, the practical sizes are 25 mm and 32 mm MDPE to the blue potable specification, and the choice between them is the whole of the pipe sizing decision on a domestic job. Twenty-five millimetres suits an ordinary house with one or two bathrooms. Thirty-two starts to earn its place on a long run, on a property with three or more bathrooms, on anything with a bath that is expected to fill quickly, and wherever the service is going under a driveway that nobody will open twice. Both are cheap relative to the excavation, and the excavation is the cost — which is the argument for going up a size while the ground is open rather than being right to the millimetre.
The fixture-unit route below gets you to a defensible bore quickly, and it is worth being clear about what it is. It is a probability method of the Hunter family, which is the basis of the tables in the International Plumbing Code and the Uniform Plumbing Code; BS EN 806-3 does the same job through loading units and a different curve, and BS 8558 carries the design flow rates for individual fittings that sit underneath it. Use the calculator to place the bore between two sizes and to see the velocity, then confirm against whichever of those documents your work is being signed off under. They will rarely disagree by more than a size on a domestic job, and when they do, the larger answer is the one that survives the next bathroom.
Resist the instinct to go much larger than the calculation asks. An oversized service holds more water for longer between draws, and stored cold water sitting in a shallow trench through a hot August arrives at the kitchen tap warm enough to taste and warm enough to matter for microbial growth. Depth of cover, ducting under hard standing, marker tape, and the sleeve and seal where the pipe enters the building are all part of the same specification — as is the rule that a plastic service should run in one unjointed length from the boundary to the internal stop tap wherever the ground allows, because the joint you cannot reach is the leak you cannot find.
The bore that carries the diversified load at a sane velocity is a separate question from what the old pipe happens to be, and it is the one that decides whether a trench removes the problem entirely.
The sum of water supply fixture unit values for every fixture served by this pipe segment.
A calibration constant that scales the square-root approximation to your building's fixture mix.
The maximum velocity allowed in the pipe to limit noise and water hammer risk.
The pressure available where the supply enters the building.
The length of pipe from the point of supply to the furthest fixture, following the run.
How much extra length is added to represent the elbows, tees and valves in the run.
The height of the highest fixture above the point of supply.
The pressure the metering assembly loses at your design flow.
The flow pressure the furthest fixture needs at its inlet to work properly.
Which Hazen-Williams roughness coefficient the friction calculation uses.
Minimum pipe diameter
0.862 in
Velocity governs this run. The diversified demand would fit through a smaller bore on pressure alone, but sustained velocity above the ceiling erodes the pipe and is audible through the building, so the velocity limit is the binding one and the pressure budget has room to spare.
- Estimated peak demand
- 14.31 GPM
- Diameter set by the velocity ceiling
- 0.86 in
- Diameter set by the pressure budget
- 0.77 in
- Pressure lost to static lift
- 58,286.8 Pa
- Pressure lost to friction at this diameter
- 122,120.36 Pa
- Pressure left at the furthest fixture
- 198,803.83 Pa
- Equivalent length used for friction
- 147 ft
They open the calculator with your figures already in it
Domestic Water Pipe Sizing by Fixture Unit Calculator: 0.8616 in — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 0.862 in — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- Fixture-unit methods work because fixtures are used intermittently and rarely together. That diversity is the whole basis of the sizing, and it is why the pipe is far smaller than the sum of the fixture flows would suggest.
- Hunter's curve is old, and its age biases this in one direction. It was derived in the 1940s from fixtures using several times the water modern ones do, and it is well documented as oversizing systems built with low-flow fittings — the demand factor is the handle for that, and lowering it is a judgement with consequences rather than a correction.
- Excludes continuous-demand loads such as irrigation and hose bibs, which do not benefit from diversity and are added at their full flow.
- The friction calculation assumes one diameter for the whole run. A real system steps down as branches leave it, so the true loss is somewhere between this figure and the loss of a system sized entirely at the smallest branch — this is the screening answer, and a segment-by-segment calculation is the design one.
- Fitting losses are an allowance, not a count. Each elbow, tee and valve has its own equivalent length and a push-fit insert reduces the bore as well, so the percentage is a stand-in for a takeoff nobody has done. On a run with many tight bends it will understate the loss.
- The meter figure is yours to supply and this page cannot check it. Meter loss rises steeply as flow approaches the meter's rating and a backflow preventer or pressure-reducing valve often costs more than the meter itself; entering a nominal figure where the real assembly loses far more is the quiet way this calculation goes wrong.
- Velocity limits govern independently of flow: sustained velocities above roughly 2.4 m/s (8 ft/s) cause erosion and noise regardless of what the fixture units allow.
- The result is an internal diameter, not a pipe size. Nominal designations are not bores — copper, PEX and CPVC of the same nominal size have materially different internal diameters, and a push-fit system loses more at every joint — so take the next size up whose actual bore meets this figure in the material you are using.
What the meter will pass, and what the undertaker will permit
A meter is an obstruction with a published loss curve, and on a marginal supply it is not a negligible one. Domestic meters are commonly older and smaller than the service they sit in, and a mechanical meter forced up near its overload flow both loses head and wears out early. The metrological classification in the BS EN ISO 4064 series describes a meter by its permanent flow Q3 and the ratio of that to its minimum flow Q1; size against the continuous rating rather than the short-duration overload, and read the loss off the manufacturer's curve for the specific model rather than assuming it is small. If the meter is in a boundary chamber and the service is being renewed anyway, that is the moment to ask for it to be reviewed.
Then read the rulebook before drawing anything. In England and Wales the Water Supply (Water Fittings) Regulations 1999 make certain work notifiable to the water undertaker before it starts, and the installation of a pump or booster drawing more than twelve litres a minute connected directly or indirectly to a supply pipe is on that list. Scotland works to the Water Supply (Water Fittings) (Scotland) Byelaws 2014 and Northern Ireland to its own 2009 regulations, to the same effect. The regulations exist because a pump on a shared main takes flow from the neighbours, and an undertaker is entitled to refuse direct boosting outright or to cap the rate at which it may draw. Get the answer in writing, at the design stage, from the undertaker whose main it is.
The same conversation should collect four other things: the guaranteed minimum static pressure at the boundary and the hour and season it applies to, the meter size and its location, whether the street main is programmed for renewal or pressure management in the next few years, and what fluid category the premises will be treated as for backflow protection. Pressure management is the one that catches designers out — an undertaker reducing main pressure across a district to cut leakage is doing exactly the right thing for the network, and it can quietly remove the margin a design was built on eighteen months after commissioning. North American readers should substitute their own adopting code and AWWA Manual M22 here; the arithmetic is identical and only the paperwork changes.
The peak the building will actually ask for is the number the meter has to pass without becoming the restriction, and it is worth having before the undertaker is asked whether the existing one is adequate.
The sum of water supply fixture unit values for every fixture served by this meter.
A calibration constant that scales the square-root approximation to your building's fixture mix.
Estimated peak demand
22.6 GPM
A meter is sized on peak demand and on the pressure it costs you, not on the size of the pipe it sits in. Fitting a meter one size up because the main is large is a common and expensive error — an oversized meter under-registers low flows.
They open the calculator with your figures already in it
Water Meter Sizing Calculator (Fixture Units): 22.63 GPM — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 22.6 GPM — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- Does not compute the meter's pressure loss at the design flow, which must be subtracted from the available pressure before sizing the distribution.
- Excludes fire demand entirely. Where a supply serves both domestic use and a fire system, the sizing basis and often the meter type are different.
- Meter accuracy classes have a minimum registrable flow. Continuous small leaks below that threshold pass unmetered, which matters for leak detection rather than for billing.
Three answers, and the deficit each one actually fixes
Once the service pipe has been cleared of blame, there are three arrangements and they are not interchangeable. Choosing between them is a matter of naming which deficit the property has. A flow deficit means the main can supply enough water over the day but not fast enough during the ten minutes that matter. A pressure deficit means the head is simply too low to reach the top floor at a usable rate whatever the flow. A combined deficit means both, which is common on high-rise conversions and on properties at the end of a long rural spur.
A booster set taken directly off the supply pipe fixes a pressure deficit and nothing else. It is the smallest, quietest and cheapest of the three, it takes no floor space worth mentioning, and it is the arrangement most likely to be refused: it adds nothing to the flow the main will give, and it can pull the neighbours' pressure down while it works. Variable-speed sets with a low-pressure cut-out are the only sensible specification, and the cut-out has to be set so the pump stops rather than starving the main.
An accumulator fixes a flow deficit and needs no electricity to do it. It is a captive-air vessel that fills slowly at whatever rate the main will give whenever the house is idle, and gives that stored water back quickly during a draw, so the property borrows against its own quiet hours. It is the elegant answer to the nine-litres-a-minute house whose main pressure is otherwise healthy, because it converts a rate problem into a volume problem — and volume is something a cupboard can hold. It also has a hard limit that gets glossed over in sales literature: an accumulator can only deliver at the pressure the main charged it to, less what friction takes on the way out. Where the main pressure is genuinely low, an unpumped accumulator gives a larger quantity of disappointing water.
A break tank with a booster set fixes anything, at a price. The tank is filled through a type AB air gap, which severs the hydraulic connection to the main entirely and satisfies the highest fluid category the premises are likely to attract, and the pump then works against a guaranteed suction rather than against the street. It is the only arrangement that is immune to what the undertaker does with pressure next year. It also brings stored drinking water into the building, which brings a Legionella regime with it, plus floor loading, a tundish and overflow that has to go somewhere visible, pump noise, a power supply and a maintenance obligation that lasts as long as the building.
The vessel sizing question is the same for an accumulator and for the tank on a boosted set, and it is asked in the same terms: how much water does the property need to draw faster than the main can refill it, and for how long. Take the design peak flow, subtract the rate the main will actually sustain at an acceptable residual pressure, and multiply the shortfall by the duration of the peak. That deficit volume is what the vessel has to hand back — and for a captive-air accumulator it is the usable fraction between the charged and the depleted pressure, not the vessel's nameplate volume, which is a considerably smaller number and shrinks further the narrower the acceptable pressure band is.
Combinations are normal and should not be treated as a failure to decide. An accumulator ahead of a small booster set is a common and good answer where the flow is short and the pressure is marginal: the vessel covers the peak, the pump covers the height, and neither is sized for the whole problem. What should be avoided is fitting the plant first and measuring afterwards, which is how a property ends up with a booster set in the garage, an accumulator in the loft and the original lead service still under the lawn.
| Arrangement | Removes | Does not help | Brings |
|---|---|---|---|
| Booster set direct off the supply pipe | Low pressure at height | A main that cannot supply the rate | Notification, possible refusal, power, low-suction protection |
| Accumulator, unpumped | Short peaks the main refills slowly | A genuinely low main pressure | Space, an annual charge check, no electrical work |
| Break tank with booster set | Both, and immunity to future pressure management | Nothing, if sized honestly | Stored-water hygiene regime, overflow, floor loading, noise |
| Accumulator ahead of a small booster | Short peaks and marginal height together | A service pipe that is the real restriction | Two items to commission and one to charge correctly |
A boosted cold supply, from the boundary stop tap up
- Boosted distribution main and risers — sized on the diversified load at a velocity the occupants will not hear, and the last chance to lose the pressure the plant just bought Domestic Water Pipe Sizing by Fixture Unit Calculator
- Pressure reducing valve — holds the boosted outlet at the figure the fittings downstream were rated for, and is selected on flow and drop rather than on the thread it lands in Pressure Reducing Valve (PRV) Sizing Calculator
- Accumulator vessel — charges slowly during the idle hours and gives the stored volume back during the peak, so a slow main can still serve a fast draw Well Pressure Tank Sizing Calculator
- Duty and standby booster pumps — variable speed against a set pressure, with low-suction protection that stops the pump rather than starving the main or running it dry
- Break tank and type AB air gap — severs the hydraulic path back to the main and gives the pump a guaranteed suction, at the cost of a stored-water hygiene regime
- Meter, check valve and strainer — an obstruction with a published loss curve, and on a marginal supply the loss belongs in the budget rather than in the margin Water Meter Sizing Calculator (Fixture Units)
- Underground service and boundary stop tap — the undertaker owns the street side of the stop tap and the property owner the rest, and this is where most of the missing pressure is buried Domestic Water Pipe Friction Loss Calculator (Hazen-Williams)
The arithmetic that sizes a pressure tank behind a pump is the arithmetic that sizes an accumulator, so enter the shortfall as the flow and the length of the peak as the run time. It applies a typical one-third usable fraction and knows nothing about your charge pressure or your acceptable pressure band, which makes its answer the smallest vessel worth pricing rather than the vessel to buy — settle that on the manufacturer's drawdown chart.
Your well pump's rated flow rate.
The shortest the pump should run per cycle to avoid premature wear.
Recommended tank size
30.3 gal
Actual drawdown fraction varies by tank precharge pressure and your specific pressure switch cut-in/cut-out settings — check your tank manufacturer's drawdown chart for a precise figure.
- Required drawdown
- 10 gal
They open the calculator with your figures already in it
Well Pressure Tank Sizing Calculator: 30.3 gal — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Estimated cost — your price
This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.
What this calculation does not cover
- The 33% usable fraction is a fixed assumption, not something this calculates. Real drawdown depends on the tank's air precharge and on your pressure switch's cut-in/cut-out settings, and the usable fraction shrinks as the switch band moves higher — so a tank set up at a high pressure band delivers less than this assumes. Check the manufacturer's drawdown chart for the model and switch setting you are actually running.
- It treats the pump's rated flow as a constant. A pump's delivered flow changes with the head it is working against and rises and falls across the cut-in to cut-out band, so the flow moving through the tank over a real cycle is not the nameplate figure this uses.
- This sizes a tank against pump cycling and nothing else. It does not look at the well's yield or recovery rate, or at peak simultaneous household demand — a tank large enough to stop short-cycling can still be far too small for a low-yield well or a heavy draw, and that is a storage problem this rule does not address.
- The fixed fraction describes a captive-air tank (bladder or diaphragm). It does not apply to a galvanized air-over-water tank, whose air charge dissolves into the water and needs periodic recharging, nor to a constant-pressure or VFD system, which is deliberately paired with a small tank and is not sized this way.
- This is not an installation specification. It says nothing about the tank's pressure rating against the relief valve setting, the stand or floor loading, freeze protection, the tank tee and pipework, or the well and plumbing code requirements that govern the installation.
Giving the house back a pressure it was built for
Boosting solves the complaint upstairs and frequently creates a new one at ground level. A set holding four bar at the manifold delivers four bar to the washing machine hose, the flexible tap connectors and the ceramic cartridges, all of which have manufacturers' ratings that are lower than people assume and fail in ways that are expensive when nobody is home. A pressure reducing valve on the boosted outlet is not an optional refinement; it is the thing that keeps the rest of the installation inside the pressure it was specified for.
Select the valve on flow and on the drop it is being asked to take, which is a valve coefficient calculation, and not on the pipe size it happens to thread into — an oversized valve at low flow hunts and chatters, and a valve asked for too large a ratio between inlet and outlet will cavitate and be audible through the structure. BS EN 1567 covers the requirements and tests for water pressure reducing valves and combination valves; every reputable manufacturer publishes a flow-versus-drop curve for each body size and that curve, not the connection, is the selection document. Leave the valve accessible with isolation either side and a gauge downstream, because a PRV drifting out of adjustment is one of the commonest quiet faults in a boosted system and there is no way to see it without a gauge.
Two consequences follow immediately from fitting one. A PRV with a check valve, or a meter check valve upstream, turns the cold distribution into a closed system, so heated water can no longer expand back towards the main; anywhere there is an unvented cylinder or a sealed heat source, the expansion provision has to be reviewed rather than assumed to be unchanged. And a boosted system with a pump that stops abruptly is a better generator of surge than the gravity supply it replaced, so check valve selection, pump ramp-down and a look at the surge arithmetic belong in the same afternoon as the valve selection.
The drop the valve has to take at the flow the house actually draws is what picks the body size, and getting it wrong is heard before it is measured.
The design (peak) flow rate the valve must pass.
The target pressure drop across the valve at design flow.
Required flow coefficient (Cv)
11.2 Cv
This is the standard simplified Cv equation for water service — always confirm the selected valve's actual rated Cv from the manufacturer's catalog at your specific operating conditions before final selection.
They open the calculator with your figures already in it
Pressure Reducing Valve (PRV) Sizing Calculator: 11.18 Cv — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 11.2 Cv — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- Cv answers flow, and what destroys pressure reducing valves is cavitation. Whether the valve survives its own drop is a separate check on the RATIO of the drop to the inlet pressure above vapor pressure, judged against that valve's own pressure recovery coefficient - a globe-style PRV taking most of its inlet pressure across one seat will flash and collapse bubbles on the downstream side, chewing out the trim and making a noise like gravel in the pipe. Large reductions are split across two valves in series or given anti-cavitation trim, and the Cv from this page will look entirely correct on the valve that fails.
- Sized at one flow, and buildings run at all of them. A direct-acting PRV droops in outlet pressure as flow rises and creeps back up as flow falls away, so a valve chosen for peak demand sits almost shut at three in the morning and can let downstream pressure drift toward inlet at no flow. The standard answer on a riser is a parallel pair - a small valve carrying night and trickle flows, a large one that opens at peak - which is a decision a single-point Cv never puts in front of you.
- A PRV turns everything downstream into a closed system. Water heated behind a valve that will not pass flow backward expands with nowhere to go, and the pressure climbs quickly enough to lift the relief valve on the heater, which is why a thermal expansion tank on the cold feed is required wherever a PRV, check valve or backflow preventer closes the system. That tank is sized from the heater volume and the pressure settings, and has nothing to do with Cv.
Commissioning, and the call that comes back in August
Commission against the measurements the job started with, not against the equipment schedule. The question being answered is whether the same three test points now read what the design said they would, at the same worst-case hour, and that is a comparison rather than a check.
- Re-measure static and dynamic pressure with flow at the boundary stop tap, so the undertaker's contribution is on record before and after.
- Prove the design peak at the worst outlet with the second-worst outlet also running, since single-outlet tests pass on installations that fail in use.
- Set and record the accumulator charge with the water side drained to atmosphere, using a gauge that reads accurately at that range rather than a tyre gauge.
- Function-test low-suction and dry-run protection by closing the suction rather than by confirming the device is wired in.
- Run duty and standby pumps separately, confirm changeover works, and check running current on each leg against the motor rating.
- Verify the break tank float valve shuts fully, that the air gap is intact and unobstructed, and that the warning and overflow discharge somewhere a person will notice.
- Set the pressure reducing valve at flow, not at standstill, and record inlet and outlet pressures at the design draw.
- Leave a written record of every reading, the vessel charge figure, and the date, in the plant space rather than only with the customer.
What the installation owes the building afterwards
Stored water changes the property's obligations. Where a break tank has been introduced, the cold water it holds is subject to the same control regime as any other stored water: the HSE Approved Code of Practice L8 and the guidance in HSG274 Part 2 set out the expectation that cold water is kept below 20 degrees Celsius and that the stored volume turns over rather than stratifying. That has a design consequence people miss — a tank sized generously for peak flow is a tank whose contents sit longer, so the sizing that looked prudent is the sizing that creates the problem. Insulate it, keep it out of the loft where it will bake, fit a screened and filtered vent, and size it for the deficit rather than for reassurance.
Materials and approvals are worth confirming rather than assuming, since boosting work introduces components that never existed on the original installation. Fittings in contact with water for domestic use are expected to meet the requirements of the Water Supply (Water Fittings) Regulations 1999 and are normally evidenced by a WRAS approval or an equivalent accepted scheme; materials in contact with water intended for human consumption also fall under Regulation 31 of the Water Supply (Water Quality) Regulations 2016 in England. Pump electrical work sits under BS 7671, and the pump itself under BS EN 809 for common safety requirements. Vessel selection should be checked against BS EN 13831 for closed expansion vessels with a built-in diaphragm, and against the Pressure Equipment (Safety) Regulations 2016 where the size and pressure bring it into scope.
Noise is the reason a technically successful job gets a complaint. A booster set bolted to a party wall transmits structure-borne noise into the neighbouring bedroom regardless of its published sound power, and flexible connections either side of the pump plus anti-vibration mounts and a plinth that is not tied to the wall are cheaper before commissioning than after. The same applies to the plant space door and to any pipework clipped rigidly to plasterboard on the run away from the pump.
Finally, leave the customer with the two facts that will keep the system honest. The accumulator's charge pressure should be checked annually with the system drained, because a vessel losing its charge quietly gives back less each year until the original complaint returns and gets blamed on the main. And a pump that starts and stops rapidly, or runs when nothing is open, is reporting a fault — a passing check valve, a dripping outlet, a lost charge — and should not be lived with. Those two habits catch nearly every developing fault on a boosted supply while it is still an adjustment rather than a replacement.
What to have measured before anything is quoted
Every line below is a reading or a written answer, not an assumption. The equipment decision falls out of them; taken in the other order it is a guess with a price attached.
- Static and dynamic pressure at the boundary stop tap — Both taken at the worst hour of the worst season, with the dynamic reading paired to a measured flow.
- Flow at the boundary, the internal stop tap and the kitchen cold — Graduated container and a watch; the differences between the three locate the restriction without any digging.
- Length, bore and material of the underground service — Confirmed by a test hole where the age of the property makes lead or part-renewal plausible.
- Design peak flow and its duration — The simultaneous draw the property has to survive, and for how many minutes — the pair that sizes any vessel.
- Meter size, location and published loss at design flow — Read from the manufacturer's curve for that model against its continuous rather than overload rating.
- The undertaker's answer on boosting, in writing — Whether a pump may draw from the supply pipe, at what rate, and what has to be notified before work starts.
Opens the calculators above on one screen with the dimensions from this article already filled in. Quantities only — this site publishes no price list, because local prices vary too much to publish honestly.