Energy Saving Calculation for Gas and Electric Usage 2

[mdalle]

I am the Resource Advisor for 17 facilities, I monitor the utility usage for these buildings. I have the gas and electric usage for the last three years for each individual building. I wanted to calcuate the savings for the last 18 months since I've implented some saving techniques. I don't think just comparing months will due, because of cooling and heating day degree difference. If the month of June was an average of 82 degrees for 2003, and a average of 95 degrees for 2004, just comparing the two will not work. Does anyone know how to calculate heating and cooling days into a converstion for energy savings?

 

[vzeos]

You should do your analysis on a degree-day basis. You can calculate the consumption (or savings) per degree-day which would put your comparison on a common footing. See the attached link for an explanation of degree-days.

http://www.vesma.com/ddd/whatthey.htm

 

[KiwiMace]

All you need to do is divide. The results are expressed in kWh per degree day or BTU/degree day etc. This normalises your results to consistent amount of demand. In a similar manner you will also be normalising for sq ft served between your different buildings.

 

[mdalle]

What do I divide. If I want to compare April 2005 to April 2006

Example April 2005 had 216 HDD(Heating Degree Days) and 41 CDD (Cooling Degree Days), whereas April 2006 had 134 HDD and 55 CDD.

For Building 1, my consumption for the month of April 2005 is 67,440 and my consumption for the month of April 2006 is 53,880.

What numbers do I use in this calculation to compare April 2005 with April 2006?

 

[KiwiMace]

you need to split out the cooling and heating energy consumption. April may have both depending on where you are.

but for example if the kWh listed was just cooling energy, your normalised answers would be:

April 2005 67440/41 = 1645 kWh/CDD
April 2006 53880/55 = 979 kWh/CDD

This is obviously not accurate, as your heating costs more than your cooling, but it shows the final method. I'm going to come back to this later - you can solve your relative heating and cooling cost simultaneously to estimate the proportion of each.

 

[KiwiMace]

The problem in the previous post was finding the split between the heating and cooling energy consumption kWh.

OK, after some wildly assumptive maths:
If we assume that kWh is in direct proportion to the heating and cooling degree days, we have two april data sets:

216H + 41C = 67440
134H + 55C = 53880

giving H = 239 kWh/HDD 37.6%
and C = 397 kWh/CDD 62.4% our proportions.

I pause to correct my previous comment in that cooling uses more energy than heating in this scenario. Cost not involved here.

So the percentages applied to the kWh should roughly give us the proportional breakdown of the total energy:

April 2005
Heating 37.6% x 67440 = 25357
Cooling 62.4% x 67440 = 42083

April 2006
Heating 37.6% x 53880 = 20259
Cooling 62.4% x 53880 = 33621

Getting back to the original purpose of finding the normalised energy use:
April 2005
25357/216 = 117 kWh/HDD
42083/41 = 1026 kWh/CDD

April 2006
20259/134 = 151 kWh/HDD
33621/55 = 611 kWh/CDD

So comparing the two years figures, it looks like the heating energy went up a bit, while the cooling energy consumption seems to have dropped significantly.

Try this on a few other months and see if the proportionality is the same over the different months. I'd be interested to hear if this method looks valid.

Metering the heating energy separate from the cooling energy is a better way forward.

 

[mdalle]

I understand most of your emial, with the exception of the Percentages. Where did you get the 37.6% for Heating and 62.4% for Cooling.

 

[KiwiMace]

239/(239+397) = 37.6% and 397/(239+397) = 62.4%

This is the proportion that each of the heating and cooling provide to the total kWh value.

 

[KiwiMace]

Just following up...

Did you check this against any of the other months to see if the proportions held?

 

[vzeos]

CinciMace, mdalle,

Clever approach, although the use of the same proportionality factors for 2005 and 2006 may not be appropriate in this case since there were energy savings measures taken between 2005 and 2006 which may have changed the proportionality of energy used for heating and cooling.

You are correct in saying that sub-metering is the best way to get a handle on the affect of the energy savings measures. I guess the next best way would be to calculate the degree-day factor for each utilization appliance (or group of appliances; or master meter) during months that do not contain mixed usage. This would be done by calculating the heating consumption per HDD for January 2005 and January 2006. Similarly the cooling degree-day factor can be found by calculating the consumption per CDD for August 2005 and 2006. The assumption here is that the energy consumption per degree-day is a characteristic of the particular appliance (or group of appliances; or master meter), which is a pretty good assumption, especially if the appliance operation is on-off and thermostat settings are reasonably constant. Using this method, the April 2005 heating consumption would be found by multiplying the April 2005 HDDs by the 2005 HDD factor. The April 2005 cooling consumption would be found by multiplying the April 2005 CDDs by the 2005 CDD factor. Similar results can be obtained for 2006.

At this point, you should be able to generate a monthly estimate of heating and cooling loads and you should be able to do a year over year comparison of your monthly energy savings.

 

[KiwiMace]

Yes, my method somewhat relied on the cooling to heating energy ratio remaining constant, which i didn't really clarify looking back, but I was hoping mdalle would follow up and check the proportionalities as I mentioned he should.

I'm not sure I follow your method given that the heating and cooling were bundled together.

Where you say...
"This would be done by calculating the heating consumption per HDD for January 2005 and January 2006"
...Do you actually mean total energy divided by HDD or CDD?

And won't the HDD/CDD factors suffer the same proportionality problem after improvements?

 

[vzeos]

"I'm not sure I follow your method given that the heating and cooling were bundled together....Do you actually mean total energy divided by HDD or CDD?"

To illustrate the method, assume a house that is heated by an electric space heater and cooled by an electric air conditioner, both of which are connected to the same electric meter.

The way to seperate out the appliance useage is to observe the meter during a period of time in which only one appliance is used and observe the number of DD for the same period. In my part of the world, the month of January never requires any air conditioning. So, if you take the meter reading for January and divide it by the cumulative DDs for January the resulting quotient would be the heating unit's DD factor (kWh/HDD).

The same could be said for the cooling load. In my part of the world, the month of August never requires any space heating so an air conditioning DD factor can be determined by using the procedure outlined above for August to get the air conditioning unit's DD factor (kWh/CDD).

Now, everything being equal (thermostat setting, appliance efficiency, building insulation, occupants etc.), the DD factors will remain reasonably constant and you should be able to use them to estimate the mixed heating and cooling consumption for any other month by using the relation:

Total Mixed Load = (HDD x kWh/HDD) + (CDD x kWh/CDD)

Note, if you undertake energy savings measures, you will need to determine new DD factors to account for the changes.


"And won't the HDD/CDD factors suffer the same proportionality problem after improvements?"

The HDD/CDD ratio is not used in the calculation.

 

[KiwiMace]

Ah, vzeos you are clearly correct.

I perhaps falsely assumed that mdalle couldn't split out the heating and cooling in any month - due to, say, process cooling in winter, or hot water generation in summer etc.