Gas Pipe Sizing Calculator

The load alone does not answer it: the length does half the work, and so does everything else on the same system. Convert first, because the sizing equation works in cubic feet per hour rather than Btu per hour. At a heating value of 1,000 Btu per cubic foot, which you should confirm with your gas supplier rather than assume, 100,000 Btu/h is 100 cubic feet per hour. One straight Schedule 40 run carrying 100 cubic feet per hour of natural gas at a half inch water column drop needs 1/2 inch at 20 feet, 3/4 inch at 60 feet and 1 inch at 150 feet (the larger of the code's own table and its equation, which agree at all three). And if that furnace shares the system with a water heater and a range, the pipe upstream of the branch carries all three, which is the part a single-pipe chart cannot tell you.

Section by section, not pipe by pipe. Draw the system as a tree from the meter outward, put each appliance's nameplate input on the section that ends at it, and then every section carries its own appliance plus everything downstream of it. The length each section is sized at depends on the method: under the longest length method every section is sized at the single longest run from the point of delivery to the most remote outlet, and under the branch length method the sections on that longest run keep it while every other section uses the longest run within its own branch. Both are permitted. The example on this page is a 235,000 Btu/h house with a 57 foot longest run: the main off the meter comes out 1 inch while the furnace drop is only 3/4 inch, because the main carries everything and the drop carries one appliance.

Which length each section gets sized at. The longest length method (IFGC 402.4.1) uses one length for the whole system, the longest run from the point of delivery to the most remote outlet, so every section is sized as though it were feeding that farthest appliance. The branch length method (IFGC 402.4.2) keeps that length only for the sections that actually sit on the longest run and lets every other section use the longest run inside its own branch, which is shorter and therefore permits smaller pipe on those branches. The informative appendix that describes them calls the branch length method less conservative than the longest length method (IFGC A103.2, which is guidance rather than code). The calculator on this page shows both, side by side, so you can see exactly which sections the choice moves and defend the one you install.

Yes, and this tool does both. IFGC 402.3 authorises the sizing tables and the equations as alternatives, so either one is code-compliant on its own. Wherever the code prints a low-pressure capacity table for the material, gas and drop you are designing to (Schedule 40 steel, copper tubing and polyethylene pipe), every section is sized both ways and the LARGER size is published, with which one governed, so no size here is ever smaller than the code's own chart for the same load and length. At a drop the code prints no table for, the equation sizes it, held to no smaller than the code's own tables printed for any looser drop (a tighter drop never carries more gas through the same pipe), and above every printed drop the equation alone sizes it; each says so. The two routes do not agree perfectly, for two reasons. A table reads the row at your length or the next longer printed row, while the equation uses the exact length, so between two printed rows the equation can come out a size smaller. And the equation as printed runs a little less conservative than the tables: for Schedule 40 steel, measured across the 1,110 cells of the two published natural gas capacity tables, where no assumed heating value can enter the comparison, about 0.9 percent; about 0.7 percent for copper at 0.3 to 1.0 in. w.c. (at the 17 in. w.c. copper table the equation is the more conservative) and 0.8 percent for polyethylene. Holding the table as a floor removes both gaps instead of disclosing them, and the equation keeps its reach at any drop and any length.

Yes, by reading the code's own CSST tables rather than computing, because the sizing equations reach only pipe or tubing with smooth inside walls and corrugated tubing is not. IFGC 402.3 authorises the code's sizing tables and a listed system's own tables alike, and the code prints low-pressure CSST tables for natural gas at 0.5 in. w.c. (IFGC Table 402.4(15)), 3 in. w.c. (IFGC Table 402.4(16)), 6 in. w.c. (IFGC Table 402.4(17)) and for propane at 0.5 in. w.c. (IFGC Table 402.4(32)). The answer is an EHD, the table's flow designation, not a diameter: each section reads the row at its length or the next longer printed row and takes the smallest EHD whose printed capacity covers the load. The appendix's own retrofit example, a barbecue 40 feet from the point of delivery at 40 cubic feet per hour and half an inch of water column, comes out EHD 18 CSST here, which is the appendix's answer. At a drop the code prints no CSST table for, or on a supply below a table's printed minimum, it refuses and says so. Pick tubing whose manufacturer lists at least that EHD, and check it against the manufacturer's listed tables.

Yes. Both are smooth-walled, so the same equation sizes them, at the inside diameter the code's own tables print: copper tubing at Type K's bore, which the copper tables' own note calls the smallest of the copper tubing products (so a Type L tube is sized conservatively), and polyethylene pipe at the bore of the SDR the tables print for each size (1/2 SDR 9, 3/4 SDR 11, 1 SDR 11, 1-1/4 SDR 10, 1-1/2 SDR 11, 2 SDR 11, 3 SDR 11, 4 SDR 11). The appendix's own copper example comes out the same as the appendix prints it, section for section. Copper sizes are K and L nominal sizes, not outside diameters, and ACR tubing is labelled one size up. Polyethylene pipe is outdoors and underground only (IFGC 404.17.1), so the tool refuses it inside a building. Polyethylene TUBING is refused, because the code's natural gas tubing tables print the 0.927 inch bore as 3/4 inch and its propane tubing table prints the same bore as 1 inch.

Because the sizing equation works in cubic feet per hour and your appliances are rated in Btu per hour, and the number that converts between them is the heating value of the gas actually being delivered. The code does not publish one. The appendix says it can be obtained from the serving gas supplier, and that is the honest route. Assuming a value undersizes the pipe wherever the real gas is leaner than the assumption, so this tool refuses to assume one: it asks, and it tells you it is asking. Undiluted propane is nowhere near natural gas on this figure, which is why changing the gas clears the field rather than carrying a natural gas number across.

No. It is what most published capacity tables ASSUME, printed in their headings, and it is a design decision rather than a code requirement. What the code actually requires (IFGC 402.6) is that the pressure at every appliance inlet meets that appliance's own minimum inlet pressure. A borrowed table assumption is not a design drop: on a long run with a low supply pressure, half an inch of water column may be more than you have to give away. State the drop you are designing to, and if you also give the tool the appliance's minimum inlet pressure it checks the drop against the lower of the supply you stated and the pressure at the point of delivery: a design that would leave the appliance inlet below its minimum is refused, with the largest drop that meets it. Without the minimum it says the check is incomplete rather than implying it passed.

The one your jurisdiction adopted, which is why this tool asks. The arithmetic is the same in every book and the address is not. International Fuel Gas Code, and the International Residential Code for one- and two-family dwellings: the sizing provisions are at IFGC 402.4, published in the IRC as G2413.4 (402.4). A homeowner pulling a residential permit is inspected against the residential code's own numbering, and printing only the commercial fuel gas address to them is a right value under the wrong authority. Two further families are offered and both refuse rather than answer, because their own coefficients were never read from their own text: putting one of those labels on this equation's answer would be a guess about most of a state's permit process.