FSP0025 – Elevators Part 3 – Facility Science Podcast #25

By | November 20, 2019

Notes for FSP0025 – Elevators Part 3
This is a continuation of elevators part 1 and 2 which are #23 and #24 of this podcast. It might be helpful to listen to those first before continuing with this one.
Elevator modes
When you push the elevator buttons (either in the hall to call the elevator or inside the car to choose a floor) the elevator controller has to make decisions about how to service all of the requests generated by those button pushes. For example, if there are calls both above and below the car, should the car go up first or down first? Also, if there are multiple cars in the system, which car should be sent to service a call? The exact thing we want to happen might be different under different circumstances, but there are also a handful or a couple of handfuls of typical situations where we might want the elevator to act in certain ways. We define these typical behaviors for certain circumstances in what we call modes of operation or service modes. So I’ll describe some of those modes now. (Note: the exact definition of these terms and the exact features of each mode might vary by manufacturer or facility)
Automatic Modes
  • Automatic mode – first there’s the default or normal operating mode, which I’ve usually heard called “automatic mode” for modern automatic elevators. This is what the elevator does when there is no special condition. Generally what happens here is that the elevator (assuming it’s starting at the bottom, say in the lobby) will start going up. On the way up it will answer any hall calls for people that want to go up (a hall call is people in the hallway that want the elevator to come pick them up) and also drop people off if it passes the floor they want to be dropped off at. It will keep going up until there are no more calls above the car. So on the way up it will ignore any down calls or any up calls below the car (so it’s not going to turn around and go back down until there are no other calls above the car. When there are no more calls above the car, the elevator will start moving down, picking up anybody that wants to go down and dropping them off when it gets to their destination and passing all of the up calls. So that’s the basic default automatic mode for an elevator. In buildings with multiple elevators operating as a group each car follows the same basic algorithm but the controller uses various criteria to decide which car will service a particular call. So if there are 2 cars heading up past a floor where there is an up call, we don’t want both cars to stop at that floor, instead the controller will use its programmed criteria to decide which car will stop.
    • This particular version of automatic mode is called selective-collective operation.
    • In this terminology, the “collective” means that the elevator will remember all the calls that are made and service them when it gets to them. This is in contrast to an older concept where an elevator would only service one call and not remember any others. Basically a non-collective elevator would ignore all calls it got while it was servicing a call, which is not particularly user-friendly. Pretty much al modern elevators are collective in this sense.
    • The “selective” in selective-collective means that, while the elevator remembers all of the calls and will eventually get to them, it is selective about the order in which it services the calls. It will only service up calls when it is traveling up and only down calls while it is traveling down, regardless of the order in which the calls were made. This is in contrast to “non-selective-collective” operation which means the elevator will service calls in the order they were received regardless of the direction of travel.
    • There are some other variations of this concept, notably down-collective and up-collective or maybe more generally something like “toward-the-lobby-collective.” The idea here (and I’ll use the down-collective version to explain) is that traffic is expected to be from the lobby to an upper floor or from an upper floor to the lobby but not from a non-lobby floor to another non-lobby floor. The elevator will take people from the lobby up to their floor and it will consider all calls from upper floors to be down calls to the lobby. So, after the elevator has dropped everyone off on the way up, it will turn around and start servicing the down calls and take them all to the lobby.
    • My understanding is that the selective-collective operation that I first described is by far the most common. We choose the others (nonselective-collective or up-collective or down-collective or even some other algorithm we might invent) when it makes sense in a particular building. And the criteria used to decide what makes sense is that we want to decrease the average wait time, so we want to tailor the elevator operation to the way the people use the building.
  • Peak mode (up peak or down peak depending on expected demand) is used when we expect there to be a disproportionate demand in one particular direction. We generally talk about up peak or down peak. And to illustrate this, you can image that between say 7 am and 10 am in an office building most people will be entering the building from the street (so into the lobby or whatever) and talking the elevators up to their offices. A this time, hardly anybody will be coming down to leave the building. Similarly, between, say, 3pm and 6pm, almost everyone will be coming down from their offices to leave the building and hardly anyone will be entering the building to go to work. Knowing this it makes sense that we design the elevator algorithm to account for the expected demand to minimize wait times. That’s what peak mode does. Peak mode can be initiated on a schedule, or by a demand algorithm in the controller or even manually.
  • In up peak mode, the cars will be parked in the lobby with the doors open until they fill up
    (Determined by car weight) or for a certain amount of time. Cars will then leave the lobby and take everyone up to their floors, then return to the lobby to repeat the cycle. People that need to come down during this time might be handled in different ways, either by having one car run on normal automatic mode, or requiring people to make a round trip through the lobby, or by some prioritization algorithm.
  • Down peak mode works similarly except that cars will be sent to the top of the building, then service all down calls to the lobby. When they get to the lobby the return to the top of the building and repeat until down peak mode is over.
  • Riot mode – this is for use where there might be some type of violent or other wise dangerous or disagreeable circumstance (civil unrest or other such situation) at the street level. Riot mode will basically prevent the doors from opening at any publicly accessible floor (so lobby floors or parking areas, whatever makes sense for a particular building) so the building occupants can operate within the building but avoid problems that might be happening outside the building.
  • Load bypass mode is activated automatically when the car is carrying too much weight. “Too much” here is defined as some percentage of the maximum capacity of the elevator and the “some percentage” is defined by the elevator manufacturer or the engineer designing an elevator system. In one case, I think it was a manual for an elevator, I read that the default for that manufacturer was that 80% of maximum load would trigger load bypass mode, but the percentage was field configurable in that model. In this mode the elevator will ignore hall calls until some people get off the elevator so the idea is that the elevator won’t pick anybody else up until somebody gets off.
  • Seismic mode (earthquake emergency mode?) – seismic sensors in the building alert the controller to potential earthquake, all cars stop at the nearest floor and open the doors. (https://www.dir.ca.gov/title8/3137.html)
  • Security Recall mode – all cars are stopped first at a security recall floor so security personnel can check out the car occupants before the car continues to the final destination
Manual modes – in contrast to the automatic mode where the controller decides how the car or cars should move, there are various manual modes where a person in the car decides where the car should move.
  • The first one I’ll talk about is Independent service mode. This mode is normally activated with a key switch inside the car. Independent service mode is designed to allow an authorized operator to take control of a car for their exclusive use. In this mode, the elevator will ignore all hall calls (so it won’t leave the floor it is on to go pick someone up), it will keep its doors open until they are closed from inside the car and it will only travel to a different floor when travel is initiated from inside the car. And then when it reaches that other  floor it will just stay there with the doors open. In a situation where there are a number of elevators grouped together, independent service mode will only affect the one car in which it is activated. So if we have a group of four elevators and we put one of them into independent service mode, the other 3 will continue to operate in automatic mode as normal.
  • A kind of variation of independent service mode is medical emergency service, often called code blue mode- is used inside a hospital or other health care facility. It allows a doctor or other authorized person to take control of an elevator (usually with a special key switch) in order to respond to an urgent medical situation. When code blue service is activated at the key switch in the hallway, the controller will cancel all calls for the selected car and send it directly to the floor where the code blue call was initiated. That car will then ignore all calls except code blue calls initiated from inside the car.
  • There is also fire service mode which you might also think of as a specialized version of independent service. Fire service is a 2 phase mode. Phase 1 is initiated either by a key switch in the hallway (usually activated by a firefighter responding to an emergency situation at the building) or by a signal from a smoke detector or fire alarm system. In fire service phase 1, the car moves to the designated emergency recall floor. The emergency recall floor will be the floor where the fire recall key switch is, so when the firefighter activate fire service mode, the elevator car will come right to where they are. In some systems there is also an alternate recall floor. If the fire detection system detects a fire at the primary recall floor, the car will go to the alternate floor instead. Ok, so when the car reaches the recall floor, it will park there with its doors open and stay there until one of two things happen. Either fire service mode is cleared using the reset function (which is available at the same key switch used to activate fire service phase 1) or the emergency response personnel activate fire service phase 2 to take control of the car. Phase 2 is activated from inside the car by a second key switch on the car operating panel. Fire service phase 2 allows the user to drive the car up and down similar to independent service mode. The elevator will ignore all hall calls. The elevator will take the user to any floor, most likely bypassing any security restrictions. The doors won’t open and close automatically, but instead are opened and closed manually by holding the door open and door close buttons. In the United States, the exact operation of fire service mode for automatic elevators is defined ASME A17.1 section 2.27.3. ASME is the American Society of Mechanical Engineers and A17.1 is Safety Code for Elevators and Escalators which actually defines many (or probably all) of the elevator safety features I have talked about in this series about elevators. The same standard is used in Canada as CSA B44-07. CSA is the Canadian Standards Association.
  • Another manual mode is attendant service mode. This is what it sounds like…a mode that allows an elevator attendant to drive the elevator up and down. You might have seen an old-timey elevator in a movie or something where the guy pulls a lever back and forth to make the car go up and down to take people to their floors. I guess it’s also possible you’ve seen such a thing in real life, but I’ve only seen it in movies…and cartoons, I guess. Anyway, modern elevators replace the lever with buttons but some still allow the elevator to be controlled by an attendant. In this mode, hall calls will still register in the elevator, but the attendant will be able to decide which order to service the calls, whether to go up or down, and how long to keep the door open at each floor.
Elevator phones – elevators have phones to allow the occupants of the car to call for help. In passenger elevators, a button push should reliably initiate 2-way communication with someone that can help the person in the car. The contacted party can be inside the building or somewhere else. Since the contacted party might be in a different building, the elevator should be able to communicate to the responding party which elevator is calling them. Might also have phone connection in the hallway so rescue personnel can call into the car to make contact with car occupants that need help. Phone requirements are defined in the same standard as fire service mode (ASME A17.1).
Traveling cables
I remember the first time I stepped into an elevator machine room (this was quite a while ago). I was working between the IT people and the elevator company to connect a phone line to the elevator. This was a hydraulic elevator, so we were on the bottom floor of the building. I remember thinking “wait a minute, we’re going to connect the phone line to this panel here, but how does it actually get to the phone in the elevator car? That thing moves way up there and comes back down.” The answer to that question is what we refer to as traveling cables…