Double Deck Examples
The Elevate files and pdf results files for these example can be downloaded here.
These examples have been prepared with Elevate Version 8.11.
This example is for an office building with 20 floors. There is a population of 130 persons per floor above the double height lobby at levels 1 and 2. Double deck elevators are being considered; there will also be escalators to link levels 1 and 2.
Up peak calculation
A simple up peak calculation using the Double Deck General Analysis allows us a quick assessment of elevator configuration is required. Assume a peak demand of 12% of the design population per 5 minutes, with the traffic 100% incoming. The analysis may be performed for a range of possible elevator configurations (number, size, speed). The target is an interval no greater than 30s and a loading by mass, and by area of no more than 80%,
Elevate file: Example DD1.elv
A second calculation has been done to assess lunch performance based on a peak demand of 13%, 45% incoming, 45% outgoing and 10% interfloor. The target is an interval of no greater than 45s and a loading by mass, and by area of no more than 80%.
Elevate file: Example DD2.elv
Based on these results, a solution of six, 3.15 m/s, 1600 kg per deck cars is selected for further study with simulation.
Using the step profile, test the performance of the system across a range of traffic intensities, from 1% to 15% with an up peak traffic flow, assuming the traffic is 100% incoming.
Elevate file: Example DD3.elv
Using the step profile we can see that the up peak performance is excellent until we reach a demand of 13% when the cars are fully loaded. This is consistent with the round trip time calculations.
Using a step profile, test the performance of the system across a range of traffic intensities, from 1% to 15% with a lunch peak traffic flow, assuming the traffic is 45% incoming, 45% outgoing and 10% interfloor.
Elevate file: Example DD4.elv
With a mixed traffic flow the system does not saturate, even at 15% demand because for each round trip, not all the people are in the car at the same time. However, the waiting times are higher.
Repeat the analysis with step profiles applying Destination Control. The waiting times are higher, but the overall time to destination is lower. The system does not saturate as people with common destinations are grouped, which reduces stops and increases handling capacity.
In this analysis the lunchtime performance with destination control is similar to conventional control.
Elevate file: Example DD5.elv
Elevate file: Example DD6.elv
The step profile is a useful tool to understand the behaviour of the system under varying loads. Now consider a more realistic representation of traffic, as described by the Peters Research (CIBSE) modern office up peak and modern office lunch peak templates.
Elevate files: Example DD7.elv, Example DD8.elv, Example DD9.elv, Example DD10.elv
The results show the conventional solution is close to saturation. The destination control solution manages the traffic; the waiting times would be acceptable in some regions. To meet the 5 star performance levels described by CIBSE, it would probably be necessary to use a seven-car group.
There is scope to examine some Elevate variables, for example, there could be improvements in door times. With wide doors, it could be argued that the passenger transfer time of 1.2s is high. With destination control, passenger transfer time is normally faster. It would be worth investigating the significance of higher speeds. With destination control, smaller car sizes may be practical. However, it is also worth considering the impact of a car being out of service.
Consider the previously considered office building has a second zone, with an additional 18 floors starting at level 21. A second group of double deck elevators express past levels 3 to 20 to serve the upper zone.
Assuming the same criteria as the previous example, an up peak calculation is performed.
Elevate file: Example DD11.elv, Example DD12.elv
Based on these results, a solution of six, 6 m/s, 1600 kg per deck cars is selected for further study with simulation in a similar manner to the single zone calculations.
Consider a building twice the height. Zone 1 and Zone 2 based on the Single Zone calculations already completed. Zones 3 and 4 are served from a sky lobby positioned between Zones 3 and 4. Passengers take a shuttle lift from levels 1 & 2 to levels 57 & 58. If they work in Zone 3, they take lifts from the sky lobby down the building. If they work in Zone 4, they take lifts from the sky lobby further up the building. This (potentially counter-intuitive) arrangement is particularly core space efficient; there are many other possible arrangements, which could be considered.
The elevators in Zones 3 and 4 would be the same specification as Zone 1, our single Zone calculations. The final task is to select the shuttle elevators.
The shuttle elevators need to transport the population of 36 floors of 130 people.
Shuttle elevators are normally sized to manage a greater handling capacity than the local elevators they serve. Historically this is because the down peak handling capacity of a conventional (non destination) elevator group is greater than its up peak handling capacity. This does not apply to shuttle elevators, which have the same handling capacity in both directions. Assume a peak demand of 15% in one direction. An up peak calculation is performed using the General Analysis. Note that the loading/unloading time per passenger has been reduced on the basis the passengers all loading two at a time.
Elevate file: Example DD13.elv
Based on this analysis, a solution of eight, 7 m/s, 2000 kg per deck cars is selected for further study with simulation.
Elevate file: Example DD14.elv
A step profile is used to investigate the point at which the selected configuration saturates. The performance in simulation is very good with saturation reached at >15% passenger demand. This is probably because the round trip time inefficiencies applied in the General Analysis calculation are pessimistic; the shuttle elevators spend a greater proportion of their time travelling at full speed than a regular elevator and suffer less from issues such as bunching.
Based on the simulation, it would be worth considering less or smaller elevators.