# Talk:Gas pipe sizing

Jump to navigation Jump to search

## Alternative Methods?

That way of sizing, although clearly expressed, is somewhat complex - therefore time-consuming and error-prone.

(a) This is a linear problem. Therefore, it should be possible to model it with corresponding electrical components, suitably scaled in value, and breadboard it. For the purpose, the components would be labelled in gas-pipe terms, so that 3m of 22mm pipe would be represented by a resistor labelled "3m of 22mm" and the corresponding length for other standard pipe diameters. The atmosphere would be represented by a ground plate full of holes into which the appliances would be mounted. Kits of parts, and accessories, would be sold to those with suitable proof of competence. It would be fun to use.

(b) More reasonably, the problem could be done by computer. To minimise the chance of human error, start with a verified HTML copy of the standard - the d-i-y FAQ page that we are discussing is sufficiently similar to that - but with the Design Data table given an ID. Then make a diagram by specifying all of the parts and connections in text, but maybe with drag'n'drop positioning. Now program the calculation exactly as in the text; the code could be interspersed with the text, with buttons to control the visibility of each. Test, and sell. Of course, that is computationally inefficient; but for any reasonably small gas system that would not matter. The Gas Boards, or whatever we have now, might need something better.

I used a similar method, maybe two decades ago, for the Date of Gregorian Easter Sunday as originally defined (Christopher Clavius, around 1600, Opera Mathematica V, in black-letter Latin, with the authority of Pope Gregory XIII); I reproduced his Tables in HTML, and followed his instructions literally by code. Testing over the repeat interval of 5,700,000 years, I found complete agreement with a method derived from the (British) Calendar Act of 1751. Which is particularly mice, because Gauss, when first reproducing Clavius's method in current algebra, did not get it quite right.

(c) If there is already something more convenient that using the Article, I suggest that it should be mentioned, independently of whether it is available for D-I-Y and/or other small users.

John Stockton (talk) 11:01, 19 May 2020 (UTC)

Hi John, yes the method described, while easy for simple installs, is potentially more complicated that it needs to be. However since it follows the standards specced way it seemed prudent to go with that since it will then make sense when combined with other sources of information. For more intensive design there are already applications available (including phone based). This a problem that could also possibly be solved in a spreadsheet with some goal seeking / What If? analysis. --John Rumm (talk) 16:08, 19 May 2020 (UTC)