A Comparison of Measure Avoided Cost Calculations

1
A Comparison of Measure Avoided Cost Calculations

• using Utility TOU Load Shapes and DEER Hourly
  Measure Savings
• 14 March 2006

2
Measure Avoided Cost Calculations

• For this analysis, we compared annual avoided
  costs for selected measures determined using IOU
  load shapes and using DEER hourly savings
  results.
• This work does NOT examine or compare the
  magnitude of measure savings in the IOU filings
  vs. DEER savings for the same measure this work
  only shows the relative avoided cost values when
  the SAME measure annual savings has an avoided
  cost calculated using the DEER hourly profile, or
  the IOU hourly or TOU profiles.

3
Measure Avoided Cost Calculations

• The components used for the analysis include
• cpucAvoided26.xls from E3
• levelized hourly avoided cost values were
  extracted from this spreadsheet for each
  utility/climate zone combination examined and for
  a range of measure expected life values.
• contains hourly TOU period definitions for each
  utility
• Utility specific shape viewer spreadsheets
• SCE-res2.xls SCE residential spreadsheet
  provides TOU values by end-use and climate zone
  for new construction, and by end-use for
  retrofit.
• SCE-nonres2.xls SCE non-residential
  spreadsheet provided TOU values by building type,
  end-use/measure-type and climate zone for new
  construction, and by building type and end-use
  for retrofit.
• SDGE3.xls SDGE spreadsheet provides TOU
  values by building type and end-use for both
  residential and commercial building types.
• ComViewer.xls and ResViewer.xls PGE
  spreadsheets that provided TOU load shapes by
  end-use for building sector. These spreadsheets
  also contain the hourly load profiles by building
  type and end-use.
• Attachment II-T3 tables
• For each of the utilities, these tables map the
  TOU load shapes used for each of the measures.

4
Measure Avoided Cost Calculations

• The DEER version of eQUEST was used for this
  analysis, with minor modifications.
• hourly whole-building electricity use for both
  the base case run and the measure run is now
  written to a file. End-use hourly values are not
  yet available.
• some improvements were made to the DEER process
  to fix known problems
• minor changes to schedules (open-closed hours)
• night-cycle control added for hot climates
• residential HVAC performance curves updated

5
Measure Avoided Cost Calculations
Measures by Building Type Measures by Building Type Measures by Building Type Measures by Building Type
Large Office Small Retail Grocery Single Family
Chiller efficiency Pkg AC efficiency Pkg AC efficiency AC efficiency
Indoor Lighting reduction Indoor Lighting reduction Indoor Lighting reduction Refg Charge
Economizer Maintenance Reduce Over-Ventilation Reduce Over-Ventilation Duct Sealing
Side Daylighting Top Daylighting Top Daylighting Low-E Window
LowSC Window on West LowSC Window on West LowSC Window on West Indoor Lighting
LowSC Window on East LowSC Window on East LowSC Window on East Ceiling Insulation
Supply Fan Motor Eff. LowSC Window on South Zero Heat Doors
CHW Loop Motor Eff. Night Covers
2-Speed Cooling Tower
Climate Zones by Utility Climate Zones by Utility Climate Zones by Utility
PGE SCE SDGE
CZ03 CZ09 CZ07
CZ13 CZ15 CZ15
6
Preliminary conclusions

• Use of TOU load shapes typically under-predicts
  the avoided cost of Commercial measures relative
  to hourly load shapes by up to 20 (or more)
• For the commercial building cases examined, the
  average difference between TOU avoided cost to
  Hourly Avoided cost is 12

7
Preliminary conclusions

• Use of TOU load shapes typically under-predicts
  the avoided cost of Commercial measures relative
  to hourly load shapes by up to 20 (or more)
• For the commercial building cases examined, the
  average difference between TOU avoided cost to
  Hourly Avoided cost is 12

8
Preliminary conclusions

• Use of TOU load shapes typically under-predicts
  the avoided cost of Commercial measures relative
  to hourly load shapes by up to 20 (or more)
• For the commercial building cases examined, the
  average difference between TOU avoided cost to
  Hourly Avoided cost is 12

9
Preliminary conclusions

• Use of TOU load shapes typically under-predicts
  the avoided cost of Residential measures relative
  to hourly load shapes by up to 30
• For the residential cases examined, the average
  difference between TOU avoided cost to Hourly
  Avoided cost is 19

10
Preliminary conclusions
The agreement between annual avoided costs
calculated with hourly and with TOU load shapes
varies by building type and measure
Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost
Measures Large Office Small Retail Grocery Single-Family
Lighting 2.7 6.3 2.9 3.5
HVAC 6.8 17.5 21.1 20.2
All Measures 6.5 14.5 37.3 20.5
11
Preliminary conclusions

• Use of a Super-TOU load shape for the DEER TOU
  Avoided Cost calculation improved the agreement
  between the hourly and the TOU annual avoided
  costs by about 20
• For the cases examined, the average difference
  between Super-TOU avoided cost to Hourly Avoided
  cost is 11 (compared to 13 for the standard
  TOU).
• For commercial buildings, the use of a super TOU
  period improved the comparison of Lighting vs.
  HVAC measures only slightly (about 10).
• For the single-family residential building, the
  use of a super TOU period did bring the Lighting
  and HVAC ratios closer together (by about 25).
• For this analysis, Super-TOU is defined as a new
  TOU period containing the 100 highest levelized
  avoided cost hours in the Summer On-Peak period.

12
Preliminary conclusions
Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost Difference between DEER Hourly and TOU Avoided Cost
Measures Large Office Small Retail Grocery Single-Family
Lighting 2.7 6.3 2.9 3.5
HVAC 6.8 17.5 21.1 20.2
All Measures 6.5 14.5 37.3 20.5
Difference between DEER Hourly and Super TOU Avoided Cost Difference between DEER Hourly and Super TOU Avoided Cost Difference between DEER Hourly and Super TOU Avoided Cost Difference between DEER Hourly and Super TOU Avoided Cost Difference between DEER Hourly and Super TOU Avoided Cost
Measures Large Office Small Retail Grocery Single-Family
Lighting 2.3 6.1 2.9 3.8
HVAC 5.7 14.7 17.0 16.5
All Measures 5.4 12.4 35.8 15.4
13
Preliminary conclusions

• Use of IOU load shapes can under-predict and
  over-predict the avoided cost of measures
  relative to DEER hourly load shapes by up to 30
  (or more)
• For the cases examined, the average difference
  between the Utility TOU avoided cost to DEER
  Hourly Avoided cost is 20.

14
Preliminary conclusions

• Use of IOU load shapes can under-predict and
  over-predict the avoided cost of measures
  relative to DEER hourly load shapes by up to 30
  (or more)
• For the cases examined, the average difference
  between the Utility TOU avoided cost to DEER
  Hourly Avoided cost is 20.

15
Preliminary conclusions

• Use of IOU load shapes can under-predict and
  over-predict the avoided cost of measures
  relative to DEER hourly load shapes by up to 30
  (or more)
• For the cases examined, the average difference
  between the Utility TOU avoided cost to DEER
  Hourly Avoided cost is 20.

16
Preliminary conclusions

• Use of IOU load shapes can under-predict and
  over-predict the avoided cost of measures
  relative to DEER hourly load shapes by up to 30
  (or more)
• For the cases examined, the average difference
  between the Utility TOU avoided cost to DEER
  Hourly Avoided cost is 20.

17
Preliminary conclusions

• Use of the IOU load shapes under-predicts the
  avoided cost for HVAC measures by a wider margin
  than lighting measures
• For the cases examined, the average difference
  between the Utility TOU avoided cost and the DEER
  Hourly Avoided cost is 9 for the indoor lighting
  measure.
• All other measure categories had significantly
  higher differences between Utility TOU and DEER
  Hourly avoided costs.

Measure Category Average Difference
Indoor Lighting Reduction 9
Cooling Efficiency Improvement 27
Economizer / OA Reduction 53
Daylighting 17
Low SC Glass on West 13
Low SC Glass on East 19
18
Preliminary conclusions

• The agreement between annual avoided costs
  calculated with hourly and with Utility TOU load
  shapes varies by building type and measure.

19
Preliminary conclusions

• Building and end use load shapes, as used by the
  current E3 calculators, do not well represent
  measure impact shapes in most cases
• The use of older building and end use load shapes
  used for the current filings amplify this
  difference.

20
Preliminary conclusions

• IOU load shapes can work quite well

when the IOU load shape follows the same trends
as the measure load shape.
21
Preliminary conclusions

• But if the TOU load shape doesnt match the
  Measure load shape

then the annual avoided cost may be significantly
under or over-estimated.
22
Preliminary conclusions

• But if the TOU load shape doesnt match the
  Measure load shape

then the annual avoided cost may be significantly
under or over-estimated.
23
Preliminary conclusions

• And even a TOU load shape that is quite different
  from the measure load shape

can lead to a correct answer due to
compensating errors.
24
Preliminary conclusions

• Some measures do not lend themselves to be
  summarized by any TOU end-use load shape

Measures that save energy during some TOU periods
but use more energy during other TOU periods
cannot be approximated by an end-use load shape.
This measure saves over 5000 kWh per year in a
typical sized grocery store, but increases demand
during the summer on-peak period.
25
Hourly Load Shape Comparison
The hourly impact of economizer maintenance has
little in common with the hourly cooling profile.
26
Hourly Load Shape Comparison
The impact of low shading coefficient glass on
the East has a much different load shape than
cooling in general.
27
Hourly Load Shape Comparison
The impact of low shading coefficient glass on
the West has a much different load shape than
cooling in general.
28
demand definition