In short. Not every load runs at once or at full rating. Maximum demand is the highest load the system actually sees. Demand factors and diversity factors estimate it from connected load. Using them well avoids oversizing; using them badly causes overloads.
Definitions
Diversity factor = Sum of individual maximum demands ÷ Maximum demand of the whole group
Demand factor is less than or equal to 1. Diversity factor is greater than or equal to 1. The first reflects that a load runs below its rating; the second that different loads peak at different times.
Typical use
- Group loads by type: motors, lighting, process, HVAC, IT.
- Apply a demand factor to each group.
- Apply diversity between groups.
- Check against measured data where possible.
- Add allowance for planned growth.
Illustrative factors
| Group | Connected kW | Demand factor | Demand kW |
|---|---|---|---|
| Process motors | 400 | 0.75 | 300 |
| HVAC | 150 | 0.80 | 120 |
| Lighting and small power | 80 | 0.60 | 48 |
| IT and controls | 40 | 0.90 | 36 |
| Total | 670 | 504 |
The factors above are examples only. Real values come from the process, measurements and applicable codes.
Applying diversity between groups
If the groups do not peak together, the maximum demand of the whole may be lower than the sum. Suppose a diversity factor of 1.1 between the groups above: 504 ÷ 1.1 ≈ 458 kW. Apply diversity cautiously, particularly for generators, where the worst-case coincident load matters during outages when everything restarts.
Standby and outage conditions
During an outage and restart, loads may all demand power at once. Diversity that holds in normal operation may not hold after a blackout. For generator sizing, consider the restart condition and the loads actually served. See restoration and restart sequencing.
Using measured data
- Record load over at least a representative period, including the busiest.
- Use the highest sustained demand, such as a 15 or 30-minute maximum, and consider short peaks separately.
- Check seasonal variation and planned changes.
Why not all loads run together
Connected load adds every rating, but loads run on schedules and cycles. Demand factor and diversity factor express this.
Demand factor = maximum demand ÷ connected load; Diversity factor = sum of individual maximum demands ÷ coincident maximum demand
Typical demand factors (indicative only, confirm by measurement)
| Load type | Demand factor range |
|---|---|
| Lighting | 0.7 to 1.0 |
| Air conditioning | 0.6 to 0.9, season dependent |
| Motors, process | 0.5 to 0.85 |
| Lifts | 0.3 to 0.6 |
| Standby equipment | 0 |
Worked example (illustrative)
Connected load 900 kVA. Applying group demand factors gives 630 kVA. Diversity between groups of 1.1 reduces the coincident demand to 573 kVA. Design demand with 15 percent growth = 660 kVA.
Where factors mislead
- Generator-backed loads may be a different subset from the whole building.
- Peak events such as restart after an outage raise demand above normal diversity because everything wants to run at once.
- Factors copied from a different building type.
Better data
Log demand for a representative week at the incomer, in 15 minute intervals, and use the measured maximum. See connected, demand and maximum demand.
Common mistakes
- Applying a demand factor twice.
- Using diversity for loads that start together after an outage.
- Reading a utility bill’s maximum demand as the generator’s design load without checking what it includes.
- Using factors from one site type for another.
What to send us
Send your load list and any demand records. Send your application details. See also connected load, demand and maximum demand.
Questions engineers ask
What is the difference between demand factor and diversity factor?
Demand factor compares maximum demand to connected load for one group. Diversity factor compares the sum of group maxima to the maximum of the combined group.
Can I use diversity for generator sizing?
Carefully. Outage restart can bring loads on together, so test the assumption for that case.
Where do typical factors come from?
Design codes, handbooks and experience. They are starting points; measured data is better.
Should I add growth after diversity?
Yes. Apply growth to the diversified demand, with a clear assumption.
Figures are illustrative. Use measured data and applicable codes for design.