There is no single acceptable lift waiting time. A figure that earns an excellent grade in a single-tenant office is a comfortable pass in an apartment block and a different question entirely in a hospital, because each building type is judged against its own benchmark, in its own traffic peak, using its own metric. This guide sets out the typical benchmark for each building type, explains why they differ, and shows how to avoid the most common mistake in applying them: holding the wrong metric up against the wrong target.
The short answer
The typical guideline figures summarised in AdSimulo’s lift performance criteria run as follows, all measured over a five-minute peak:
- Offices, morning up-peak: an average interval of 25 seconds or less is graded excellent for a mixed-tenancy building and 20 seconds or less for a single-tenancy one. Group handling capacity of 12% or more of the population is the usual floor for mixed or diversified tenancy, and 15% or more for a single tenant.
- Offices, lunchtime two-way peak: an average waiting time of 40 seconds or less, with group handling capacity of 12% or more.
- Hotels, evening two-way peak: an average interval of 40 to 45 seconds for a commercial hotel and 40 to 50 seconds for conference and tourist hotels, with handling capacity of 10 to 12% or more.
- Residential apartments, evening two-way peak: an average interval of 45 to 50 seconds for luxury apartments, 50 to 60 seconds for normal apartments and 50 to 70 seconds for low-income housing, with handling capacity of 8% or more for luxury apartments, falling to 5 to 7% for low-income housing.
- Car parks, one-way peak: an average interval of 40 to 45 seconds for an office-related car park and 45 to 60 seconds for a free-standing one, with handling capacity of 8 to 10%.
- Hospitals: no single benchmark, for reasons covered below.
Two details sit behind those numbers and change how they should be read. Most of them are intervals, not waiting times. And every one of them belongs to a specific traffic peak, so a figure is only meaningful against the peak it was set for.
Interval is not waiting time, and the difference matters
Average interval is the gap between consecutive lift arrivals at the main lobby. Average waiting time is how long a passenger actually stands there between registering a call and a car arriving. They move together, but they are different measurements, and a benchmark written in one cannot simply be read as the other.
The office guidance shows why this matters. For lifts with conventional control, the guideline is an average interval of 30 seconds or less. For lifts with destination control, the guideline is an average waiting time of 25 seconds or less, and the British Council for Offices Guide to Specification 2014 continues to recommend an average waiting time of less than 25 seconds. Quote “25 seconds” without saying which metric, and which control system, and the number means nothing. The hotel, residential and car park figures above are all stated as intervals, so a 35-second average waiting time from a hotel study cannot be checked directly against them.
The practical rule is simple: before comparing a result to a target, confirm that both are the same metric. The difference between waiting time and the whole journey, which matters most in tall buildings, is covered in AdSimulo’s guide to waiting time versus time to destination, and how to read each metric on a report is covered in how to read a lift traffic analysis report.
Why benchmarks differ by building type
Three things drive the differences.
The first is what occupants will tolerate. People arriving at a prime office in a tight morning window have little patience for a lobby queue, and in a commercial building lift service is part of what a tenant is paying for. Residents of an apartment block spread their journeys across a longer evening period, and the published benchmark intervals for apartments are correspondingly longer.
The second is the shape of the peak. An office up-peak is a concentrated one-way surge from the main lobby. A hotel or apartment evening peak is a two-way flow spread over a longer period. A car park peak is one-way and driven by arrival and departure patterns. A target set for one shape does not transfer to another.
The third is the grade of the building. Every benchmark above comes in bands, excellent, good, fair and unacceptable for offices, luxury, normal and low income for apartments. Picking the right band is a decision to agree with the client.
Each benchmark also pairs two measures: a quality measure, how long people wait, and a quantity measure, how much of the population the lifts can move in five minutes. A design has to meet both: a group with a good interval but weak handling capacity is describing lifts that arrive promptly and still cannot clear the queue.
Offices: the tightest benchmarks, and why tenancy matters
Office benchmarks depend on tenancy, because tenancy changes both the population and how sharply it arrives. The guidance defines three types: diversified tenancy, where no single tenant occupies more than a single floor and no more than 25% of tenants share the same business activity; mixed tenancy, where a single tenant has multi-floor occupancy or several tenants share the same activity; and single tenancy, where one tenant occupies 80% or more of the building. The typical population densities are 14 square metres of net internal area per person for diversified tenancy, 12 for mixed and 10 for single.
Having calculated the total population, an absentee factor of 15 to 20% is normally applied for leave, sickness and meetings, and a further discount may apply where stairs are easily used at the lowest floors. The result is the working population every other figure depends on, and it is the input to confirm with the client before anything else.
For the morning up-peak, average interval is graded as follows:
- Excellent: 25 seconds or less for mixed tenancy, 20 seconds or less for single tenancy.
- Good: 30 seconds or less for mixed tenancy, 25 seconds or less for single tenancy.
- Fair: 35 seconds or less for mixed tenancy, 30 seconds or less for single tenancy.
- Unacceptable: more than 35 seconds for mixed tenancy, more than 30 seconds for single tenancy.
Group handling capacity, as a percentage of the zone population in five minutes, is graded by tenancy type:
- Excellent: over 14% for diversified tenancy, over 15% for mixed and over 16% for single.
- Good: 13 to 14%, 14 to 15% and 15 to 16% respectively.
- Fair: 11 to 12%, 12 to 13% and 13 to 14% respectively.
- Unacceptable: under 11%, under 12% and under 13% respectively.
The guideline floors, according to the BCO 2009 and CIBSE Guide D guidelines, are 15% or more for a single-tenant building and 12% or more for mixed or diversified tenancy. The BCO Guide to Specification 2014 simplifies this to an up-peak handling capacity of at least 12%, and specifies a traffic mix of 85% up, 10% down and 5% inter-floor for the up-peak. That mix matters as much as the target, because a result is only comparable if it was produced under the same demand.
Offices also have a second peak. For the lunchtime two-way peak, the recommendations are an average waiting time of 40 seconds or less and a group handling capacity of 12% or more, and the 2014 guide specifies a lunchtime demand of 13% with a 45% up, 45% down and 10% inter-floor split. In buildings with internal restaurants, shared facilities across levels or mixed uses, this peak can be harder to serve than the morning one, which is the case made in AdSimulo’s guide to interfloor traffic.
A common rule of thumb is one passenger lift for every 250 to 300 office users. It is useful as a sanity check and no more, since the benchmarks above are what actually decide the number.
Hotels: evening two-way peak and the population per room
Hotels are judged on the evening two-way peak, and the benchmark depends on the type of hotel because the type changes the population. Typical figures are:
- Commercial hotels: 1.3 to 1.7 persons per room, an average interval of 40 to 45 seconds or less, and handling capacity of 10 to 12% or more.
- Conference hotels: 1.8 to 2.0 persons per room, an average interval of 40 to 50 seconds or less, and handling capacity of 10 to 12% or more.
- Tourist hotels: 2.0 to 3.0 persons per room, an average interval of 40 to 50 seconds or less, and handling capacity of 10 to 12% or more.
As a rule of thumb, one guest lift is required for every 100 guest rooms. One caveat matters more than the table: most major hotel operators publish their own guidelines for both guest and service lifts, and where an operator standard exists, it replaces the generic figure. Hotels also load differently from offices. Luggage means a hotel car reaches its effective limit well before its rated headcount, which is why the load factor assumed for an office should not be borrowed.
Residential apartments: the most tolerant benchmarks
Residential buildings are also judged on the evening two-way peak, with the most tolerant intervals of any building type:
- Luxury apartments: an average interval of 45 to 50 seconds or less, with handling capacity of 8% or more.
- Normal apartments: 50 to 60 seconds or less, with 6 to 8% or more.
- Low-income apartments: 50 to 70 seconds or less, with 5 to 7% or more.
The population is built from the apartment type and the housing category. Occupancy factors, in persons per apartment for luxury, normal and low-income housing, are 1.0, 1.5 and 2.0 for a studio; 1.5, 1.8 and 2.0 for a one-bedroom; 2.0, 3.0 and 4.0 for a two-bedroom; and 3.0, 4.0 and 5.0 for a three-bedroom. The specifics of tall residential schemes are covered in the guide to lift traffic analysis for residential towers.
Car parks: one-way peak, driven by arrivals per vehicle
For car parks, the population is estimated from the number of parking spaces and the average number of people per vehicle, with a discount for users who prefer the stairs at the lowest floors. The typical figures are:
- Office-related car parks: 1.2 to 1.5 persons per vehicle, an average interval of 40 to 45 seconds or less, and handling capacity of 8 to 10%.
- Free-standing car parks: 2.0 to 2.5 persons per vehicle, an average interval of 45 to 60 seconds or less, and handling capacity of 8 to 10%.
Other types, such as retail-related car parks, need the two-way peak analysed instead.
Hospitals and mixed-use buildings: no single number
A hospital cannot be given one benchmark because it has no single governing peak. Bed and trolley movement, staff shift changes, visitor flows and continuous 24-hour demand overlap throughout the day, and a lift group has to be tested against the combination rather than one condition. The approach is set out in the guide to lift traffic analysis for hospitals, and the sensible course is to agree the target with the client, then test it against every peak that matters.
Mixed-use buildings raise a related problem. ISO 8100-32, the international standard for the planning and selection of passenger lifts in office, hotel and residential buildings, applies to mixed-use buildings provided each use can be evaluated separately as office, residential or hotel. The overlap between those uses is where a combined study earns its place, as shown in AdSimulo’s guide to mixed-use lift simulation.
How to test a design against a benchmark
A benchmark is only useful if the comparison is like for like. A reliable sequence is:
- Establish the working population. Use the tenancy, the apartment mix or the hotel type, and confirm the figure with the client.
- Choose the building type and the grade. Confirm whether an operator or client standard replaces the generic one.
- Match the metric. If the target is an interval, compare an interval. If it is a waiting time, compare a waiting time.
- Run the peaks that apply. That means the up-peak for an office, and the lunchtime two-way peak as well, or the evening two-way peak for a hotel or apartment block.
- Look at the distribution as well as the average. An average can hide a bad five minutes, which is where simulation adds what a calculation cannot.
For a first check, AdSimulo’s lift traffic calculator returns average interval and handling capacity for an idealised up-peak with every assumption shown. For a full study, AdSimulo’s traffic analysis software tests the building as drawn against the targets you set, and its Executive Scorecard grades the design against those targets so a shortfall is flagged rather than buried.
Common mistakes when applying benchmarks
- Comparing an interval to a waiting-time target, or the reverse, so the result looks better or worse than it is.
- Using an office benchmark for a hotel or apartment block. The peak, the population and the tolerance are all different.
- Testing the up-peak only in a building whose lifts are sized by the lunchtime or evening peak.
- Treating a band as a pass or fail. Excellent, good and fair are different promises to a tenant.
- Ignoring an operator standard. Hotel brands often publish their own, and it governs.
- Quoting a figure without its peak and its traffic mix. An 85/10/5 up-peak and a 45/45/10 lunchtime peak give different answers for the same lifts.
Benchmarks are guidance, not law
None of the figures in this guide is a statutory limit. CIBSE Guide D, published through the CIBSE knowledge portal, provides guidance rather than a standard, and the BCO figures are specification targets for prime offices, not legal requirements. They matter because they are the shared reference a client, a planning authority or a peer reviewer will measure a design against. The role of Guide D in that process is explained in the guide to CIBSE Guide D.
The full tables behind this guide, including the office, hotel, residential and car park figures, are in the lift performance criteria reference. Once the right benchmark is chosen, the work that remains is testing a real building against it, which is what a traffic study is for.










