Daylight Savings Time Impact

1
Daylight Savings TimeImpact
2
DST Impact

• Goal - estimate the change in energy due to DST
• Approach Measure energy before and after the
  DST date

3
DST Impact

• Two days were used for comparison
• Selected one day before the DST change and one
  after
• Attempted to make sure that the weather was as
  close as possible
• Chose the same day of the week
• Used 560 residential sample accounts from GPC
• Compared the five morning hours and five evening
  hours
• Any potential DST impact would occur the during
  5am - 9am and 7pm - 11pm hours

4
DST Impact

• Chose April 3 and April 17, 2003
• Both days were Thursday
• Similar hourly temperature

morning 3-Apr 17-Apr evening 3-Apr 17-Apr
5am 52 59 7pm 71 66
6 51 58 8 68 66
7 51 58 9 66 65
8 55 58 10 64 65
9 60 60 11 63 64
average 54 59 average 66 65
5
DST Impact

• kW demand between the two days was not
  significantly different

kW 3-Apr 17-Apr
morning 10.1 9.9
evening 11.4 11.7

• Paired t-test was used to statistically determine
  if the difference in usage was significant
• t-test results showed no change in usage after
  the DST shift

6
DST Impact

• Repeated the analysis with two random days
• Chose April 1 and April 8, 2004

mean temp 1-Apr 8-Apr
morning 39 58
evening 48 67

• April 1 was a much colder day than April 8

7
DST Impact

• Unlike the first test, there was a significant
  difference in usage between the two days

mean kW 1-Apr 8-Apr
morning 10.1 7.4
evening 9.9 8.2

• April 1, which was much colder, had a higher kW
  usage than April 8
• This process was repeated with other days,
  yielding similar results

8
DST Impact

• Conclusions
• When temperatures are somewhat similar, there is
  no significant difference in usage
• Temperature has an impact on demand
• DST does not affect demand

9
DST Impact

• A separate analysis was done using Alabama
  residential data
• ANOVA test (similar to paired t-test) was used
• Looked at each day for the entire week before and
  after DST
• Examined the two hours before/after sunrise
• Identical conclusions were reached
• Temperature is the key driver of demand
• DST does not influence demand