Galvanized Pipe Flow: Two Exponents, and Why the Thin Layer Beats the Rough Surface

Two Mechanisms, and One of Them Is Nearly Invisible

An aged galvanized line loses capacity in two separate ways at the same time, and the one that looks trivial does as much damage as the one that looks catastrophic.

The first is the pipe surface. The zinc coating roughens, corrodes and tuberculates, and the Hazen-Williams C factor falls with it. Published references give galvanized steel about 120 when new and 70 or lower after twenty years of service, which reads as a total collapse of the interior surface, because it is one.

The second is the bore. Scale and corrosion product grow inward from the wall, and the opening water passes through gets smaller. A tenth of an inch of that layer is thin enough that a plumber cutting the pipe open would call it light scale and think nothing of it.

Head loss varies inversely with the C factor raised to the power 1.852 and inversely with the internal diameter raised to the power 4.87. Run those two exponents on a one inch Schedule 40 line and the collapse of the surface across two decades multiplies head loss by 2.71, while the light scale multiplies it by 2.80. The invisible mechanism wins, and the whole of the difference is in the exponents.

What the Calculation Is Actually For

The Hazen-Williams relation needs a diameter and a C factor. On a new pipe both are lookups. On a sixty year old service line neither one is known, and that is the entire problem.

The nominal size is a trade designation and describes nothing physical. The schedule bore is what the pipe had when it left the mill. Neither describes the pipe in front of an engineer standing over an open trench, and the general pipe flow calculation assumes both are already in hand. This calculation exists to establish those two inputs, state which route produced each of them, and hand them across.

That framing sets a limit worth stating early. No source consulted publishes a remaining bore against age for galvanized water pipe, because the rate depends on water chemistry, temperature, flow history and the individual line. There is therefore no age based bore estimate here and there should not be one anywhere: an estimate with no published basis is a guess wearing a number, and it would be raised to the power 4.87 on its way into the answer.

What the calculation does instead is separate the two mechanisms, report each with its exponent named, and make the sensitivity of the result to the bore visible rather than burying it. The arithmetic itself is on the calculator page and is not repeated here as the substance of the article.

Calculator Inputs: Scope, Bore, C Factor, Capacity

Twenty fields in five groups, of which the first group is not hydraulic at all.

Unit System. US customary (ft, in, gpm) or SI (m, mm, L/min). Every quantity is held in inches, feet and gallons per minute internally and converted at the edges, so the toggle changes the presentation and never the physical result.

Application. Water service line, interior domestic water distribution, fire sprinkler piping, or other. This field decides whether a federal replacement category can apply at all, and it is established before anything is applied.

Lead Service Line History. Whether the line is or ever was downstream of lead, is downstream of a lead status unknown line, is upstream of any lead piping, was demonstrably never downstream of lead, or is unknown. On a service line this field, not the flow figure, produces the consequential answer.

Nominal Pipe Size. 1/2 in (DN 15) through 6 in (DN 150), Schedule 40 steel. The schedule bore comes from ASME B36.10M outside diameter less twice the wall.

How the Remaining Bore Is Known. A measured bore from a cut section or camera survey, a scale thickness sensitivity at a stated thickness, or the schedule bore used deliberately as a clean pipe comparison. Every figure that depends on the bore states which of the three produced it.

Measured Bore [in or mm] or Scale Thickness Per Wall [in or mm]. One or the other, depending on the route above. The second field is labelled per wall everywhere it appears, for a reason the arithmetic section returns to.

C Factor Source. A field flow test, a project design basis, a code or standard mandate, an estimate, or the published band. A C entered without a source is flagged, and a source without a value is incomplete.

C Factor Value and Approximate Age of the Pipe [years]. The first where a C is entered directly, the second where the published band is used, in which case the age selects a figure from the band rather than computing one.

Water Condition Over the Pipe's Life. Soft or aggressive, moderate, hard or scale forming, treated municipal supply with corrosion control, or unknown. Context for the reading, not a coefficient.

Pipe Length [ft or m], Design Flow [gpm or L/min], Available Head [ft or m]. All optional. Without a length the answer is a set of ratios, which hold whatever the length is. With a length and either a flow or a head, the ratios become figures in feet or metres.

Evidence. A cut section, a camera survey, visibly reduced flow, discoloured water on first draw, or none. What the reader has actually seen, which is what decides how much of the answer is a measurement and how much is a what-if.

The relations behind the outputs:

h_ratio  = (C_new / C_now)^1.852 × (D_new / D_now)^4.87
q_ratio  = (C_now / C_new) × (D_now / D_new)^2.63
D_now    = D_new − 2 × t_scale
h_f      = 10.44 × L × Q^1.852 / (C^1.852 × D^4.87)
Q_head   = [ h_f × C^1.852 × D^4.87 / (10.44 × L) ]^0.54

The constants that are not fields:

C_new = 120, the published C factor for galvanized
          steel when new
1.852 = the exponent on the C factor in Hazen-Williams
4.87  = the exponent on the internal diameter for
          head loss at a fixed flow
2.63  = the exponent on the internal diameter for
          flow at a fixed head loss
10.44 = the US customary Hazen-Williams constant,
          feet, gallons per minute and inches

What is missing from the field list is as informative as what is in it. There is no field that turns an age into a bore. There is no coefficient that turns a water condition into a scale thickness. There is no single aged C factor, because the published figures disagree with each other at the same age. And there is nothing about fittings or valves, because the equivalent length of the route belongs to a different calculation.

The Bore Loses Twice the Scale Thickness

Scale grows on every part of the wall, so a layer of a given thickness closes the bore by twice that thickness. This is the single most consequential arithmetic detail on the page.

D_now = D_new − 2 × t_scale

D_new    schedule bore when new, in
t_scale  scale thickness PER WALL, in
           0.05 to 0.20 in typical of the range worth
           testing on a small service line

A tenth of an inch per wall on a one inch Schedule 40 line:

D_new = 1.049 in
D_now = 1.049 − 2 × 0.10 = 0.849 in
lost  = 0.200 in, which is 19.1 % of the original bore

Enter a total diameter loss into a per wall field and the calculation subtracts it once instead of twice. The bore comes out at 0.949 inches instead of 0.849, the head loss multiplier at 1.63 instead of 2.80, and nothing about the answer looks wrong. That is the failure mode worth guarding against, because a plausible number attracts no scrutiny. The field is labelled per wall on the input, in the result row, and in the sensitivity table beneath it, three times for one quantity, deliberately.

The same thickness is not the same problem on every size. A fixed 0.10 inches per wall takes 32 percent of the bore of a 1/2 inch line and 3.3 percent of a 6 inch one, and after the exponent that is the difference between an emergency and a rounding error:

1/2 in   0.622 → 0.422 in    head loss × 6.61
3/4 in   0.824 → 0.624 in    head loss × 3.87
1 in     1.049 → 0.849 in    head loss × 2.80
2 in     2.067 → 1.867 in    head loss × 1.64
4 in     4.026 → 3.826 in    head loss × 1.28
6 in     6.065 → 5.865 in    head loss × 1.18

Which is why the complaint arrives from houses on small service lines and almost never from buildings fed by a four inch main of the same age and the same water.

The C Factor Is a Band, Because the Sources Disagree

Four published statements about galvanized steel are in general use, and they do not agree at the same age. Reporting a single aged C would assert a precision none of them supports.

• About 120 for new galvanized steel. This is the baseline every multiplier on the page is measured against.
• 120 falling to about 100 after a few years, as the zinc surface roughens.
• 120 degrading to 70 or lower after twenty years of service, described in the source as a critical consideration for asset management and rehabilitation.
• 100 mandated by NFPA 13 Table 23.4.2 for unlined and galvanized steel in fire sprinkler work.

The last of those is the one that gets misused. It is a mandated design value inside one application, conservative by intent and accounting for decades of service tuberculation. It is not a measurement of aged galvanized pipe in general, and carrying it into a domestic water calculation imports a sprinkler design convention into a place it was never meant to go. The calculator names it as a sprinkler figure and keeps it visually apart from the three observational statements.

What the band costs in head loss, with 120 as the reference:

C = 120   head loss × 1.00   (new)
C = 100   head loss × 1.40
C =  70   head loss × 2.71

Two things follow. First, the whole published range of surface degradation, from new to badly tuberculated across twenty years, is worth a factor of 2.71. Second, and this is the part worth carrying away, the step from 120 to 100 is worth only 1.40, and 0.035 inches of scale per wall on a one inch line does the same damage. Thirty five thousandths of an inch. Nobody would call that scale at all.

The Sensitivity Table Is the Honest Form of an Unknown Bore

Where the bore has not been measured, the page runs a sensitivity rather than an estimate, and shows the entered thickness alongside its neighbours so the shape of the curve is visible:

0.05 in/wall   bore 0.949 in   head loss × 1.63   flow 77 %
0.10 in/wall   bore 0.849 in   head loss × 2.80   flow 57 %
0.15 in/wall   bore 0.749 in   head loss × 5.16   flow 41 %
0.20 in/wall   bore 0.649 in   head loss × 10.36  flow 28 %

Each 0.05 inch increment costs more than the one before it. The first costs 0.63 in the multiplier, the last costs 5.20. That acceleration is what the exponent of 4.87 means physically, compounded by the fact that the bore closes from both sides at once.

The same curve sets the value of a measurement. A ten percent error in the bore moves head loss by a factor of 1.67, so a bore known to within ten percent still leaves head loss uncertain by two thirds of itself. That is the argument for cutting a section or running a camera: the measured bore is the only figure on the page that is not an assumption, and every other number inherits its uncertainty amplified by nearly five orders of exponent.

It is also why the sensitivity is labelled as a what-if on every dependent figure. The result answers what the capacity would be if that thickness of scale exists. It does not say that it does.

The Evidence Decides How Much of the Answer Is a Measurement

Two engineers can enter identical numbers and hold results of completely different standing, and the difference is what either of them has actually seen.

A cut section is the strongest evidence available and produces the only figure on the page that is not an assumption. It also produces something the arithmetic cannot: a look at whether the deposit is a smooth uniform layer or hard nodular tuberculation, which is the difference between a bore that is genuinely circular at the reduced diameter and a bore that has an effective flow area smaller than the smallest measurement across it.

A camera survey gives the same class of answer for a longer run and catches the thing a single cut section cannot, which is variation along the line. Scale is not uniform over 150 feet. A survey that finds the worst section is worth more to a head loss calculation than an average would be, because head loss accumulates along the whole length while the worst section sets the local velocity.

Visibly reduced flow over time is a symptom rather than a quantity. It confirms the mechanism is active and gives no bore. It is worth anchoring against one figure though: on the worked one inch line, flow at the same head halves at a bore of 0.806 inches, which is 0.12 inches of scale per wall. An occupant reporting that the shower is half what it was is describing something in that region, not something marginal.

Discoloured water on first draw points at corrosion product mobilising from the pipe wall, which is consistent with an interior in the state the low end of the C band describes. On a service line it also touches the other half of the page, because the mechanism behind the federal category is precisely that a galvanized line holds and releases particulate matter it collected from upstream.

No direct evidence is a legitimate entry and the page reports it as one. Everything downstream of the bore then reads as a sensitivity, which is the honest description of a calculation whose most influential input has not been observed.

Water Chemistry Explains the Spread Without Supplying a Number

The obvious next question is why two lines of the same age in the same city can differ by a factor, and the answer is the water that ran through them.

Soft, aggressive water attacks the zinc and then the steel beneath it, tending toward corrosion and tuberculation rather than a smooth deposit. Hard, scale forming water lays down mineral deposit on top of whatever the surface has become. Treated municipal supply with corrosion control chemistry, orthophosphate typically, deliberately builds a protective film and slows both processes. Temperature accelerates all of them, which is why the hot side of a building of a given age is routinely worse than the cold.

What none of that supplies is a coefficient. No source consulted maps a water condition onto a scale thickness or a C factor, and the honest reason is that the mapping would have to carry decades of flow history, temperature profile and treatment changes to mean anything. The field exists on the calculator as context for the reading and it deliberately does not enter the arithmetic.

Where it does earn its place is in choosing which sensitivity to run. A sixty year old line in soft water with no corrosion control history is a case for testing the upper end of the range and finding out what happens at 0.15 or 0.20 inches per wall. A twenty year old line in a system with a long orthophosphate programme is a case for the lower end. That is a judgement about which what-if is worth looking at, which is a very different thing from a number the calculation could produce on its own.

Length Cancels, Which Is Why the Ratios Come First

The page answers in ratios before it answers in feet, and that ordering is a property of the relation rather than a presentation choice.

Head loss in Hazen-Williams is proportional to length. Take the ratio of the aged pipe to the same pipe when new and the length appears in both the numerator and the denominator, so it cancels:

h_now / h_new = (C_new / C_now)^1.852 × (D_new / D_now)^4.87

The multiplier of 7.60 and the flow ratio of 33 percent therefore hold over 40 feet and over 400. They are statements about the pipe rather than about the installation, which is why the calculator will return them with no length entered at all, and why they are the two figures worth carrying into a conversation with somebody who has not measured anything yet.

Turning them into figures takes a length and then one of two further inputs, and each combination answers a different question:

length + design flow      how much head the line now
                            consumes at the flow it was
                            meant to carry
length + available head   how much flow the line can
                            actually pass with the head
                            the site has
length + both             the margin, positive or
                            negative, at the design flow

Entering a flow or a head with no length gets its own report rather than a number, because there is nothing to run the friction over. That is worth knowing before the fields are filled in: the length is the cheap input here, usually available from a site plan or a tape measure, and it is what converts a ratio nobody can act on into a head loss figure somebody can take to a decision.

A one inch Schedule 40 galvanized service line, 150 feet long, roughly twenty years old, in a system that can demonstrate the line was never downstream of lead. The bore has not been measured, so the sensitivity is run at 0.10 inches per wall. The C factor is taken from the published band at that age, which gives 70.

The two mechanisms, separately:

C effect     (120 / 70)^1.852       = 2.71 × head loss
bore effect  (1.049 / 0.849)^4.87   = 2.80 × head loss

They compound rather than competing, because both are terms in the same denominator:

combined  = 2.71 × 2.80  = 7.60 × head loss
flow      = (70/120) × (0.849/1.049)^2.63 = 0.33

Seven and a half times the head loss, and a third of the flow the pipe delivered when it was new. Now turn the ratios into figures. At a design flow of 10 gpm over the 150 feet, with 20 feet of head available:

head loss at 10 gpm   94.6 ft now,   12.4 ft when new
flow at 20 ft of head  4.3 gpm now,  12.9 gpm when new
head margin           74.6 ft short now, 7.6 ft to spare when new

The line was adequate with a margin of 7.6 feet when it was installed and is now 74.6 feet short of its own design flow. Nothing about the installation changed. The pipe is in the same trench at the same length feeding the same fixtures, and the only difference is a rough surface and a layer of scale thin enough to be called light.

In metric, entering 2.5 mm per wall on the same DN 25 line, the bore goes from 26.6 mm to 21.6 mm, 5 mm lost, and the combined multiplier comes out at 7.47. The small difference from 7.60 is only that 2.5 mm is slightly less than a tenth of an inch, which is a useful check that the conversion is happening on the quantity and not on the label.

Why Lining Recovers the Smaller Half

Rehabilitation options divide cleanly along the two exponents, and the division decides what a treatment can and cannot buy.

A treatment that restores the surface without restoring the bore acts on the C factor and leaves the diameter where it is. In the worked example that addresses the factor of 2.71 and leaves the factor of 2.80 in place. Head loss falls from 7.60 times to 2.80 times, which is a real improvement and also the smaller half of what was lost.

Worse, a lining occupies part of the bore it is applied to. A coating thick enough to matter mechanically enters the calculation at the power 4.87 as a further reduction in diameter, working against the C factor improvement it was installed to deliver. On a one inch service line already down to 0.849 inches, the arithmetic of that trade is unforgiving; on a four inch main the same coating thickness is close to free. The exponent does not change, but the fraction of the bore it applies to does.

Replacement restores both terms at once, which is the only intervention that does. That is worth stating plainly rather than leaving as an inference, because the flow figures on their own tend to be read as a case for the cheaper option, and the exponents say otherwise for exactly the small diameter lines where flow complaints originate.

None of which is a recommendation. It is the arithmetic of what each option can reach, and the choice between them belongs to the engineer with the rest of the context.

On a Service Line, the Category Outranks the Flow Figure

There is a second question about an aged galvanized line that has nothing to do with hydraulics, and where it applies it is the consequential answer.

Under the federal Lead and Copper Rule Improvements, a galvanized service line is classified as galvanized requiring replacement where it is currently or ever was downstream of a lead service line, where it is currently downstream of a lead status unknown service line, or where the water system is unable to demonstrate it was never downstream of a lead service line. The mechanism behind the category is that such a line can adsorb upstream lead particulates and contribute lead to drinking water even after the original lead source has been removed. Categorisation sits in 40 CFR 141.84, with the definition in 141.2.

Where the category applies, full replacement is required, on a compliance date of 1 November 2027 and a replacement deadline ten years after it. Improving the flow does not move a line out of the category. Rehabilitation does not, a partial replacement does not, and ownership is not an exemption where the system has legal or physical access to conduct a full replacement.

Three limits on that statement matter as much as the statement:

The defined term is a service line. Interior domestic distribution piping is not one, however lead the service line feeding it once was. The calculator establishes the application before it applies the category, and interior piping is never classified as galvanized requiring replacement. Telling somebody federal law schedules their interior pipe for replacement is the most serious error a page like this could make, because a reader has nothing to check it against.

An unknown history is not a finding of lead. Where the history cannot be demonstrated, the rule treats the line as requiring replacement. That is a statement about how the rule handles missing evidence, not a statement that lead is present.

The classification is the water system's to make. It belongs to the system's own service line inventory, and a short lead connector such as a gooseneck at the main is enough on its own to put a line back into the category. A galvanized line sits outside it only where the system can determine it was never downstream of lead or of a lead connector.

Every regulatory statement on the calculator carries the date it was last verified, currently 25 August 2026, because this is live regulation rather than static engineering data. The compliance date, the replacement deadline, the action level and the categorisation rules must be checked against the current rule, the adopting state's own regulation and the water system's inventory before being relied on.

Standards and References

• 40 CFR 141.84, Lead service line inventory, replacement, and notification requirements (US Environmental Protection Agency). The service line categorisation requirement and the four categories, with the definition of a galvanized service line requiring replacement in 40 CFR 141.2.
• EPA Lead and Copper Rule Improvements, questions and answers for states and public water systems (US Environmental Protection Agency, 28 November 2023). The three limbs of the galvanized requiring replacement definition, the adsorption mechanism behind the category, the lead action level of 10 parts per billion, the national minimum average annual replacement rate of 10 percent, the prohibition on partial replacement outside emergency repair or coordinated infrastructure work, and the treatment of a line as under the system's control where legal or physical access exists.
• NFPA 13, Standard for the Installation of Sprinkler Systems, Table 23.4.2 (National Fire Protection Association). The mandated Hazen-Williams C factor of 100 for unlined and galvanized steel pipe in fire sprinkler hydraulic calculations, which is a design value inside that application rather than a measurement of aged pipe generally.
• ASME B36.10M, Welded and Seamless Wrought Steel Pipe (American Society of Mechanical Engineers). Outside diameters and wall thicknesses from which the Schedule 40 bores used here are taken, outside diameter less twice the wall.
• The Hazen-Williams relation in US customary form, head loss in feet as 10.44 × L × Q^1.852 / (C^1.852 × D^4.87) with L in feet, Q in gallons per minute and D in inches. The exponents of 1.852 on the C factor and 4.87 on the diameter are the whole argument of this article.

FAQ

Related Calculators

• Hazen-Williams Pipe Flow: the general friction calculation this page hands its bore and C factor to, taking a diameter, a C factor, a length and a flow and returning head loss and velocity (article).
• Water Service Line Sizing: where a replacement line gets sized, which is where a line in the federal category is actually going.
• Water Pipe Sizing: the interior distribution a replacement service connects to, sized against velocity and friction limits by material.
• Equivalent Length of Pipe Fittings: the developed length this calculation is run over, once the fittings and valves in the route are accounted for.
• Water Pressure Calculator: the pressure the line has to work with, against which a head loss of 94.6 feet has to be judged.
• Copper Pipe Sizing: a replacement specified in Type L copper, with its own bore table and roughness.
• PEX Pipe Sizing: a replacement specified in PEX, where the smaller bore of the CTS dimensional series matters more than the higher C factor.
• Water Meter Sizing: the meter sitting in the same service, whose own pressure loss competes for the head the aged line is consuming.
• Booster Pump Sizing: the pump whose duty a restricted line changes, and the equipment most often specified to compensate for a bore nobody has measured.