dlj wrote:
That's quite interesting. So the weather service is using a very simple algorithm at first and then adjusting later you think, is that what you are saying?
No! I'm saying they use only the high and low temperature for each day to determine that days mean and that days HDD's. They will adjust the highs and lows retrospectively to reflect actual data vs. their original forecast, and that moves the retrospective HDD bar accordingly (more often than not also retrospectively improving my consumption accuracy vs. HDD's).
It would make a lot of sense to me that over time your ability to predict based on HDD measurements - in retrospect - could become quite accurate over multiple day calculations. You are running a fairly constant inside temperature I believe. Have you tried moving that temperature up or down and see how it impacts your coal consumption? You are using the 65 baseline for the HDD calculation, but how are you using your internal house temperature set-point in your consumption calculations? At this point are you simply trying to keep that a constant and ignoring it?
dj
The 65 degree HDD 'base' is the norm for a 70 degree maintained home or office building, as those who developed the HDD system (not me) determined that heating is generally only required when external temperatures fall below that temperature. As Lightning has suggested, other heat sources apparently suffice down to 65 degrees.
For every degree I go above 70 degrees I have determined that it costs me (on average) ~3% in additional fuel consumption, and for every degree below 70 I save ~3%. But since base 65 HDD's are intended for a 70 degree maintained structure, and I'm comfortable with 70 degrees, why should I deviate from it?
I do not determine my consumption in retrospect. I determine it in advance. I go out one week, and I adjust my forward consumption forecast over that week daily as Accuweather adjusts their forecast. If the weather forecasting is accurate, the HDD's based coal consumption (or electrical consumption, etc...) is highly accurate as well. I fill the boilers hopper roughly every 3-4 days. Every 2 days only when it gets really cold. So I'm projecting future usage for generally 2 to 3 hopper fillings weekly. It often surprises me how frequently my addition of coal for any individual hopper filling hits the projected figure square on the head.
I've seen corporate level statistical studies that claim HDD's are considered about 80% accurate for heating and 89% accurate for cooling. For my case I'm getting typically at least 90% accuracy for each individual hopper loading, and about 95% accuracy over one month spans. This for November through April. October and May are skewed by the fact that my boiler needs to burn unnecessary coal to maintain a fire on warmer days when there is no home heating taking place.