August
The model overpredicts summer consumption.
← All projects
04 / Energy analysis & economics
Residential Energy Audit
Read the data. Question the answer.
A household energy model can match the annual bill and still miss the seasons. This audit connected the numbers to the decisions they should—and should not—support.
The decision
Compare the energy model with bills, then evaluate savings against replacement cost.
The evidence
12 months of consumption data; annual agreement checked alongside larger monthly errors.
A year in energy
Summer changes everything.
FEB 2025 — JAN 2026 kWh
Feb
530 607.9
Mar
479 495.6
Apr
1,067 1,035.3
May
1,521 1,391.5
Jun
1,832 1,840.7
Jul
1,880 1,903.5
Aug
1,760 2,111.2
Sep
1,408 1,684.8
Oct
1,572 1,067.5
Nov
1,011 594.1
Dec
603 559.4
Jan
502 591.3
View the underlying data
| Month | Bill-derived | Modeled | Difference |
|---|---|---|---|
| Feb | 530 | 607.9 | 14.7 % |
| Mar | 479 | 495.6 | 3.5 % |
| Apr | 1,067 | 1,035.3 | -3.0 % |
| May | 1,521 | 1,391.5 | -8.5 % |
| Jun | 1,832 | 1,840.7 | 0.5 % |
| Jul | 1,880 | 1,903.5 | 1.3 % |
| Aug | 1,760 | 2,111.2 | 20.0 % |
| Sep | 1,408 | 1,684.8 | 19.7 % |
| Oct | 1,572 | 1,067.5 | -32.1 % |
| Nov | 1,011 | 594.1 | -41.2 % |
| Dec | 603 | 559.4 | -7.2 % |
| Jan | 502 | 591.3 | 17.8 % |
14,165
Annual bill-derived consumption · kWh
13,882.8
Annual modeled consumption · kWh
−2.0%
Annual model difference, rounded
01 / Build a baseline
Start with what
the home actually uses.
The study combined a twelve-month billing baseline with an end-use model of air conditioning, lighting and appliances.
Consumption was derived from the electricity bills using the report’s assumptions: 0.18 SAR/kWh, 15% VAT and a 10 SAR monthly meter charge. This is a reconstruction of that study, not advice about today’s electricity tariff.
A degree-day regression connected consumption to weather conditions. Its reported fit, R² = 0.7124, suggests that weather explains a meaningful share of variation—but not the full picture of occupancy, habits and equipment use.
Reported degree-day fit
E = 2.6349 × DD + 362.43 E: monthly electricity, kWh · DD: report degree days 02 / Look beyond the annual total
Close over a year.
Less close month to month.
October
The actual bill remains higher than the model.
November
The largest relative mismatch needs attention.
Errors can offset across the year. A small annual difference is useful, but it does not establish a reliable month-by-month model. Better operating schedules and seasonal assumptions would be the next refinement.
03 / Follow the dominant load
Cooling sets
the summer agenda.
Air conditioning represents 59% of the modeled annual consumption—and 76% in July. That changes which interventions deserve attention first.
Annual
July
December
Air conditioning Appliances Lighting
The proposed behavior changes included a 24°C cooling setpoint, reduced unnecessary lighting time and lower standby use. These are modeled recommendations, not observed post-intervention savings.
04 / Efficiency is not the whole answer
Energy savings.
A different economic story.
The dedicated economic analysis shows why replacing working equipment cannot be justified by energy savings alone under the study’s assumptions.
| Option |
Modeled saving
kWh / year |
Investment
SAR |
Simple payback
years |
|---|---|---|---|
| Five air conditioners | 1,835.7 | 11,000 | 33.3 |
| 46 lighting units | 202.99 | 368 | 10.07 |
| Refrigerator | 66 | 1,500 | 126.3 |
Lighting proposal also reduces light output from 700 to 540 lm per lamp; it is not a like-for-like brightness comparison. Simple payback excludes discounting, maintenance and replacement timing. The dedicated economic tables are used because later summary figures conflict.
The takeaway
The right recommendation
is not always new equipment.
The audit supports starting with operating habits and better seasonal modeling, then evaluating replacements in context. It demonstrates the difference between a technical efficiency improvement and a financially attractive decision.