Home > Climate Monitoring > Societal Impacts > Residential Energy Demand Temperature Index (REDTI)

Residential Energy Demand Temperature Index (REDTI)

Overview

The Residential Energy Demand Temperature Index (REDTI) is based on population weighted* heating and cooling degree days, and as such, is a valuable tool for explaining year-to-year fluctuations in energy demand for residential heating and cooling. Residential energy consumption is known to be highly correlated with heating and cooling degree days. Energy consumption increases as the number of heating and cooling degree days increases and falls as the number of heating and cooling degree days falls. Diaz and Quayle (1980) found the correlation between energy use and heating degree days to be as high as 0.97 at the household level. Because of this strong relationship, seasonal changes in the REDTI can provide a good indication of the nation's fluctuating energy demands.

The Residential Energy Demand Temperature Index is calculated on a seasonal basis, using the sum of population weighted HDD's and CDD's (base 65), to provide retrospective information on the impact of seasonal temperatures on residential energy demand from 1895 to the present. To simplify year-to-year comparisons, the index is scaled from 0 to 100. An index of 100 is assigned to the year with the greatest population weighted degree day average while the year with the smallest degree day average receives an index of 0.

In creating the winter season index, for example, the population weighted degree day totals are calculated for each winter season. From this series, the maximum and minimum yearly degree day totals are identified. The minimum value is then subtracted from all years in the series so that a new series of values is created which ranges from zero to a value equal to the arithmetic difference between the maximum and minimum. This new series is then scaled to have a range of 0 to 100 using a common scaling factor**.

To determine how well the index captures year-to-year changes in energy demand, the REDTI was correlated with residential energy consumption*** for the period 1980-2000/01. National residential energy consumption values are available from 1973 through the winter of 2001, but because the later half of the 1970's was a period of dramatic change in energy conservation methods, high energy prices, and changing demand patterns, these years were omitted from the analysis. The effects of an increasing trend in residential energy consumption since 1980 were removed by linearly detrending the energy consumption time series prior to the correlation analysis. Because other factors, such as the effects of generally increasing US temperatures, are also removed from the detrended energy consumption time series, the REDTI was also detrended. Figures 1 through 4 contain scatter plots of detrended energy consumption and the REDTI during each season as well as the resulting correlations which were not less than 0.70 and as high as 0.86.

plot of winter energy usage versus REDTI Figure 1. Scatter plot of winter residential energy usage versus winter Residential Energy Demand Temperature Index for the period 1980-2001.
plot of spring energy usage versus REDTI Figure 2. Scatter plot of spring residential energy usage versus spring Residential Energy Demand Temperature Index for the period 1980-2000.
plot of summer energy usage versus REDTI Figure 3. Scatter plot of summer residential energy usage versus summer Residential Energy Demand Temperature Index for the period 1980-2001.
plot of fall energy usage versus REDTI Figure 4. Scatter plot of fall residential energy usage versus fall Residential Energy Demand Temperature Index for the period 1980-2000.

Table 1 contains residential energy demand in trillion BTU, and Table 2 contains the seasonal REDTI values. Both sets of data are for the period 1973 - 2001 prior to detrending. An example of the national summer season REDTI since 1895 is shown in figure 5.

Table 1. Seasonal Residential Energy Consumption (Trillion BTU)
Year Winter Spring Summer Fall
1973 -99.99 2579.18 1498.73 1951.18
1974 3628.97 2489.05 1495.99 1884.21
1975 3627.84 2726.49 1488.86 1774.54
1976 3816.05 2423.96 1518.60 2047.49
1977 4374.53 2304.37 1513.11 1858.53
1978 3953.53 2600.49 1544.55 1922.16
1979 4000.89 2474.05 1440.79 1776.35
1980 3471.58 2368.02 1427.07 1784.27
1981 3629.90 2093.52 1373.33 1718.97
1982 3536.30 2290.66 1335.45 1659.21
1983 3285.20 2252.59 1387.82 1579.67
1984 3438.48 2346.90 1401.25 1597.51
1985 3465.04 2145.44 1414.45 1654.83
1986 3505.12 2184.09 1467.53 1689.36
1987 3348.40 2238.92 1534.29 1793.17
1988 3625.51 2322.79 1581.73 1859.81
1989 3509.65 2432.66 1577.12 1899.69
1990 3605.87 2231.15 1581.31 1835.83
1991 3474.39 2230.56 1615.04 1938.16
1992 3500.18 2322.37 1557.37 1931.95
1993 3725.23 2530.14 1703.95 1992.66
1994 3981.27 2371.95 1673.96 1870.41
1995 3631.23 2398.42 1727.06 2033.93
1996 4108.81 2661.36 1747.47 2123.97
1997 3949.95 2484.42 1758.01 2113.72
1998 3708.79 2476.37 1869.42 1981.55
1999 3729.12 2550.28 1894.37 2020.56
2000 3889.10 2391.90 1913.38 2188.35
2001 4405.76 -99.99 -99.99 -99.99
Table 2. Seasonal Residential Energy Demand Temperature Index
Year Winter Spring Summer Fall
1973 49.15 37.83 46.96 18.93
1974 35.99 44.98 17.01 40.85
1975 31.74 98.74 31.73 9.23
1976 33.83 33.10 3.760 95.84
1977 95.26 19.15 58.21 37.87
1978 92.63 75.14 42.79 45.20
1979 96.25 49.19 13.08 37.34
1980 43.05 68.56 87.63 65.01
1981 46.63 40.72 56.28 41.88
1982 71.35 71.24 25.72 30.27
1983 18.70 69.04 71.86 27.48
1984 64.55 89.29 42.77 41.41
1985 56.81 21.87 30.92 28.93
1986 49.60 19.82 51.02 47.29
1987 37.11 31.99 59.49 43.67
1988 53.77 44.56 77.24 53.99
1989 41.69 56.50 27.39 33.42
1990 41.59 37.63 46.50 11.80
1991 28.38 33.61 59.44 45.98
1992 13.18 47.68 0.00 56.66
1993 47.58 63.30 65.42 60.37
1994 65.07 40.36 52.24 18.77
1995 22.50 54.37 74.40 51.28
1996 52.56 89.61 36.39 62.97
1997 28.51 64.35 23.27 57.31
1998 6.08 43.59 81.96 26.02
1999 11.13 55.24 71.83 1.87
2000 17.88 26.17 38.50 52.13
2001 65.14 51.86 -99.99 -99.99

Missing values are denoted by -99.99

Figure 5. Residential Energy Demand Temperature Index for the summer season since 1895 (National).

*Population weighting is based on 2000 census figures in 344 climate divisions throughout the period of record. The use of population weighting in averaging degree days across the nation results in a national degree day average that more closely reflects large temperature deviations in heavily populated areas of the country.

**Calculation of scaling factor: The scaling factor was calculated using the maximum value in the new series. This scaling factor was calculated to be the number that, when multiplied by the maximum value in the new series, resulted in a value of 100. The degree day totals for all years in the new series were then multiplied by the same scaling factor to arrive at the scaled series.

***Residential Energy Demand figures were supplied by the Energy Information Administration and are comprised of residential coal, natural gas, petroleum and electricity (excluding electricity losses) usage.

References

  • Diaz, H.F., and R.G. Quayle, 1980: Heating Degree Day Data Applied to Residential Heating Energy Consumption. Journal of Applied Meteorology, 3, 241-246