Part-load performance: why buildings rarely run at design

HVAC equipment is sized for design conditions: the coldest or hottest weather the building is expected to see, with full occupancy and every light on. Those conditions happen only a handful of hours a year. For the rest of the year, which is nearly all of it, the system runs at part load.

This simple fact has big consequences. The efficiency that matters most for energy cost is not the full-load rating on the nameplate but the performance across the range of partial loads where equipment actually spends its time. Systems that are efficient at design but poor at part load often disappoint. This guide explains why, and how to design for the real operating profile.

Why design conditions are rare

Design conditions are defined statistically. For example, a 1% annual cooling design value is exceeded in only about 1% of the hours in a year, and the NBC January 2.5% value is exceeded in about 2.5% of January hours. See selecting outdoor design conditions for Canadian cities.

Several other factors push actual loads further below design:

• Weather coincidence: peak temperature, peak solar and peak humidity rarely happen at the same hour.
• Occupancy: buildings are seldom fully occupied.
• Internal gains: lighting and equipment loads are lower than design assumptions much of the time.
• Safety factors and oversizing: margins added during design increase installed capacity. See diversity and safety factors in load calculations.
• Equipment size steps: selecting the next larger model adds more margin.

The result is that many systems spend most of their operating hours well below half of their installed capacity.

Load profile and bin analysis

A load profile shows how many hours a system operates at each fraction of its capacity. It can be estimated from:

1. Bin analysis: grouping annual weather hours into temperature ranges (bins) and calculating the load in each.
2. Hourly energy modelling: simulating all 8,760 hours of the year using a weather file.
3. Trend data: for existing buildings, measured operating data from the building automation system.

The profile typically shows a long tail of low-load hours and very few hours near design. Equipment selection should be informed by where the hours are, not just where the peak is.

Part-load efficiency ratings

Recognizing that full-load efficiency is a poor predictor of annual energy use, industry standards define part-load metrics.

How IPLV works

AHRI 550/590 calculates integrated part load value (IPLV) as a weighted average of efficiency at 100%, 75%, 50% and 25% load. The weighting factors are 1%, 42%, 45% and 12% respectively, reflecting that most operating hours occur at 50 to 75% load. IPLV uses standard rating conditions, including entering condenser water temperatures that fall as load falls. NPLV uses the same weighting at project-specific conditions.

IPLV is a useful comparison tool, but it represents a generic building and climate. For a specific project, a bin or hourly analysis with the actual plant configuration is more accurate. See chiller selection, efficiency, IPLV and part load.

What makes equipment efficient at part load

Variable speed

Variable-speed drives on fans, pumps and compressors allow output to match load closely. Because fan and pump power falls roughly with the cube of speed in variable systems, reducing speed saves a great deal of energy. See fan laws and how to apply them. In practice, savings are less than the ideal cube law because systems must maintain a minimum pressure.

Multiple stages and modular equipment

Several smaller units operating in sequence can match load more closely than one large unit, and they often operate closer to their best efficiency point. Modular boilers and chillers also provide redundancy.

Condensing and lower temperatures

Condensing boilers become more efficient at part load because lower firing rates and lower return water temperatures increase condensation. Chillers become more efficient when condenser water temperature falls with lower load and cooler weather. Resetting supply temperatures based on load takes advantage of these effects.

Heat exchangers with more surface

Coils and heat exchangers sized generously relative to the load have smaller temperature approaches at part load, which improves plant efficiency.

Where part load causes problems

Part-load operation also creates design and control challenges:

• Short cycling of single-stage equipment that is far larger than the load.
• Poor humidity control when cooling equipment cycles or operates at high coil temperatures.
• Control valve and damper instability when they operate nearly closed.
• Low delta T in chilled water systems when coils are overflowed at low load.
• Minimum turndown limits on boilers, chillers and VAV boxes that are reached frequently.
• Ventilation shortfalls in VAV systems where outdoor air delivery depends on supply airflow.

Each of these should be checked against the lowest expected load, not just the design load.

Designing for part load: a checklist

1. Calculate loads accurately and minimize unnecessary margin.
2. Develop a load profile, using bin analysis or energy modelling.
3. Select equipment for good efficiency across the profile, using IPLV, IEER or project-specific analysis.
4. Choose variable-speed fans, pumps and compressors where load varies significantly.
5. Divide large plants into multiple units sized to cover low loads efficiently.
6. Check turndown at the lowest expected load for every major component.
7. Specify reset strategies for supply air, chilled water, hot water and static pressure. See variable speed pumping and differential pressure setpoints.
8. Commission and trend the system across seasons to confirm part-load performance.

Frequently asked questions

Is full-load efficiency irrelevant?

No. It sets peak electrical demand and matters for facilities that run near capacity, such as some data centres. For most commercial buildings, though, part-load efficiency has more effect on annual energy use.

Why did my new high-efficiency equipment not save as much as expected?

Common reasons include oversizing, controls that do not take advantage of part-load features, missing reset strategies and operating conditions different from the rating conditions. A retro-commissioning review often identifies the cause.

Is IPLV a guaranteed annual efficiency?

No. It is a standardized comparison metric based on a generic weighting and rating conditions. Actual performance depends on the building, climate, plant configuration and controls.

Key takeaways

• Design conditions occur only a small fraction of the year; most hours are at part load.
• Load profiles from bin analysis, energy modelling or trend data show where the hours are.
• Part-load metrics such as IPLV and IEER better reflect annual efficiency than full-load ratings.
• Variable speed, staging and temperature reset improve part-load performance.
• Check turndown, stability and humidity control at minimum load, not just at design.