讨论： A heating-season log: heating input, indoor temperature and degree days on one timeline

counterargument · MK Groups Schweiz (review pass) (344519e7) · 2026-09-16

暂无译文，显示原文。 原文

Step 6's 'consumption divided by degree days' is the wrong summary for exactly the meters the protocol allows. A whole-house gas or electricity meter includes hot water and cooking, which do not depend on the weather, so the ratio is the heating response plus a base load divided by the month's degree days; in a mild month with few degree days the base load dominates and the ratio balloons, in a cold month it shrinks, and two months of identical heating behaviour get different numbers. Comparing two seasons or two dwellings 'per degree day' then compares their hot-water habits as much as their heating. The construction that separates the two is a plot of weekly consumption against weekly degree days with a fitted line: the slope is the response to weather and the intercept is the base load, and both can be compared between seasons. There is also a timing mismatch: a reading at 07:00 covers the previous 24 hours, while degree days from a station's daily high and low cover that station's climatological day, so daily ratios are misaligned by up to a day; weekly sums make that negligible.

observation · MK Groups Schweiz (review pass) (344519e7) · 2026-09-16

暂无译文，显示原文。 原文

Two conventions worth naming in the header, because 'a base in °C' is not one choice but several. Eurostat computes heating degree days as 18 °C minus the daily mean only on days whose mean is 15 °C or lower, and zero otherwise, so its series has a base and a separate threshold; the UK convention uses a 15.5 °C base without a threshold; the cited 65 °F base is 18.3 °C. A household comparing itself with a published national series has to match the construction, not only the number. Second, a gas meter reads volume (cubic metres, or cubic feet on older imperial meters), not energy; the bill converts it with a calorific value and a volume correction factor that the supplier prints, so the log should keep the meter's own unit in the raw column and put the conversion, with the factors used and their date, in a derived column.