There’s barely a day goes by where we don’t see a misplacement of the term ‘Interval’, be it in correspondence, in conversation or otherwise. Almost unbelievably, the most frequent offender is in modern specification documents: we’re talking employers requirements, brand standards and contract tender packages to name but a few. As you might imagine, text which becomes contractually binding is not a place to be mincing your words.
To understand why this might be, a quick history lesson is required. Stick with me!
Classically, lift system performance was analysed using calculation alone (prior to the development of robust computer simulation tools). The calculation used is known as the ‘up-peak round trip time’ (UPRTT) as it evaluates the round trip time of a single lift with all of the traffic in the upwards direction, for a given handling capacity. This was based on initial work by Jones (1922) and Schroeder (1955) and was formalised by Barney and dos Santos (1985) in a manner that is widely accepted by the modern lift industry. Dividing the UPRTT result by the number of lifts in the group provides us with a value for ‘Interval’.
It’s important to understand that the concept of ‘waiting time’ was not possible to evaluate using an UPRTT, or indeed any of the enhancements or improvements that were made to that calculation in the years that followed, most notable of which by Peters (Formulae for the general case, 1990). Hence, ‘Interval’ was the best available means to link system performance to passenger experience, as it describes the average time between lift cars departing the home floor. It therefore became the de-facto means of laying down a marker for lift performance in the late 1980’s and remained as such until arguably the early 10’s, in the UK at least.
There you are, history doesn’t have to be boring!
So what’s the actual problem with ‘Interval’ appearing in a performance spec in 2019? Seems like a pretty good measure of system performance, developed by some of the greatest mathematical minds the lift industry has ever seen, right? Well, yes. Precisely that – interval is a system parameter, it is not an accurate indicator of passenger experience. Using ‘Interval’ alone as a design parameter can mask poor (or even catastrophic) real world performance. Consider the following theoretical examples:
A low rise building with a single elevator and an extremely low intensity traffic flow. The round trip time of this lift could be infinite, as the system is effectively not being used. The Interval would be the same as the UPRTT, as there is only one lift in the group (Interval = UPRTT / lifts in group). If targeting a design interval of 40s, it wouldn’t be possible to demonstrate this by calculation unless more (phantom) traffic was added to the lift group to increase the car departures, or the quantity of lifts were increased. In reality passengers would experience a nominal waiting time with a single lift.
A high rise building with multiple elevators and extremely intense traffic. In this case, we need to visualise that the interval describes the frequency between loaded lift cars departing the home floor. Now imagine the morning up-peak traffic flow in an office and queues beginning to form in front of the lifts. No matter how long the queue and the associated length of waiting time, the interval would not increase past it’s maxima as full lift cars continue to depart at a given frequency. Some queueing is inevitable in modern buildings but interval alone provides no way for us as designers to evaluate this, potentially masking poor performance in reality.
It’s important to note that we generally commence our traffic studies by using a General Analysis or UPRTT calculation to test the theoretical handling capacity of the system. The lesson here is not that ‘Interval’ is wrong, but that it cannot be considered in isolation. Modern traffic analysis uses dynamic simulation tools, which are more powerful and sophisticated than the classical calculations, permitting us to evaluate lift systems on the basis of ‘average waiting time’. This allows us to interrogate passenger experience in addition to system response.
Our recommendation is that you set your requirements for lift performance in terms of ‘average waiting time’, not ‘Interval’. This minimises the risk of an overdesigned or underperforming installation. CIBSE Guide D 2015 lays down modern markers for acceptable waiting times, along with methodology for carrying out simulations to attain reliable values for average waiting time.
At Vercon we maintain an up-to date, balanced view on industry standards relating to passenger lift traffic analysis, planning and design – complemented by a strong connection to the developmental history of these practices. If you want to understand more about this topic and how it can help add value to your scheme, we love making new friends – just give us a call.