Sometimes those elevator door buttons actually work, from Henry Gifford

Once upon a time I measured how much electricity it takes to ride on an elevator – I measured about 50 elevators – and seem to be the only person who ever measured that and got the results published. Before, there were numerous scholarly papers published on elevator energy use, all guesses or estimates or computer models (same thing), very “scholarly” because they had footnotes referencing other published guesses and because you paid a lot of money for someone to do that “research”. Several multinational elevator companies were selling energy efficient elevators, but none of the companies knew how much energy a ride took – they had “data” on how much their elevator used in one year vs. an ordinary elevator – an unspecified number of rides, unspecified number of stops (floors), unspecified weight in the cab, etc. All just made up “data”. Surprise, none of the manufacturers wanted me to measure their elevator.

Along the lines of this discussion, almost everyone who heard that there was no data despite the existence of many elevators around the world for so long, insisted that the data existed, and that articles about how much energy it took to ride on an elevator had been published. Apparently it was too psychologically painful to even consider that nobody in the world knew how much electricity an elevator ride required. For some years I was offering $1,000 to anyone who could produce even one measurement not taken by me, but no takers – just lots of people insisting the data was there and the articles were there. No, nobody ever specified where “there” was, they just insisted it was somewhere but didn’t try collect the $1,000 for telling me where it was. I guess it was too painful to live in a world where nobody cared enough to measure, ever, despite so many “green” elevators being sold for so many “green” buildings around the world.

While measuring energy use I also measured elevator speed, to see if buying a slightly slower elevator saved electricity (I found no correlation). Interestingly, nobody ever seemed to wonder how in the world I measured the speed of an elevator, or was offended that I measured. Manufacturers quoted the speed at full speed, but as the elevator gradually accelerates, then runs full speed for a while (usually a short while on anything but a tall building), then slows to a stop, how could one measure the top speed during the middle of the ride? Simple. I measured how long a five stop ride took (from the door starting to close to the door being fully open on the new floor), then measured how long a six stop ride took, and knew the difference in time and the difference in distance (floor-to-floor measured with a tape measure in a stairway). It must have been not psychologically painful at all to believe that I had measured the speed, despite the speeds I measured lacking any correlation to the speeds shown in the manufacturer’s literature, as nobody doubted or objected or started shouting when I mentioned measuring speed.

I found it very interesting that my measurements of how long a ride took were almost always within 1/10 of a second of each other, thus apparently the rides were taking the same time as each other, and my measurement technique was very good (Radio Shack stopwatch on a cord around my neck). I cannot think of any other explanation for my measurements being the same for ride after ride.

While measuring energy use I investigated those buttons. When the cab gets to the new floor the doors open very soon after the car “stops” at the new floor. Stretch in the cable of a cable elevator (not in the kind lifted by a hydraulic piston) causes the cab to bounce up and down a bit as it comes to a stop, which feels like a significant delay between the cab stopping and the doors opening. In reality, it feels like the cab has stopped, but it is still bouncing around a bit too much to open the doors just yet. No, the button does nothing if pushed as the elevator is stopping at the new floor.

But if someone is seen running to catch the elevator, pushing the “open” button does open the doors on any elevator I checked, and holds them open until an annoying buzzer or bell discourages people from holding the doors open for more than a few seconds.

The more common question is if the “close” button does anything. Yes, on some elevators I checked, it does cause them to close more quickly. On others it appears to do nothing.

I found it fascinating how few elevators had buttons arranged in a vertical row in sequence of how the floors are actually arranged in the building. One parking garage had floors marked P1, P2, and P3, arranged horizontally, leaving the rider with about a 50% chance of choosing the top or bottom floor that they want to go to. The middle floor was, of course, easier to find on that elevator panel. Many other panels have few enough floors for one vertical row of buttons to be reached by a person in a wheelchair, yet arrange them horizontally, or more commonly, in columns. Many other elevators had open and close and other non-floor function buttons that looked the same as the floor buttons, increasing confusion and adding valuable seconds to the ride. No doubt some buildings could eliminate the cost and space of one elevator if the buttons were arranged so riders could push the right button a few seconds faster, but the problem is that the elevator manufacturers are set up to make the same panels they have been making for decades, and none of the people I knew who were buying elevators at the time had the courage to refuse to pay hundreds of thousands of dollars for an elevator with a control panel not arranged as they chose, thus there seems to be no solution to the problem. A minor problem for sure, but as a low-cost solution would benefit so many people a bit, it is interesting to see how hard it is to solve a problem with such an obvious solution. Maybe that is the problem – people are embarrassed to fix the problem because that would mean acknowledging that existing panels have so much room for improvement.

Humbert H. asks:

How did you measure how much energy an elevator uses?

Henry Gifford answers:

Measuring energy use is difficult, at best. Quite different than some of the measurements discussed on this list. And, like some other things discussed on this list, other important things simply cannot be known without extreme measures. And, like many other things discussed on this list, few people understand how to measure some important things, and despite ever greater computing power, other things are getting worse.

For example, a great portion – sometimes more than half - of an electrical bill and many gas bills for all but a single dwelling unit is the “demand charge” in proportion to the highest 15 or 20 minute use (“peak”) in the past year. Meters record the highest peak use in the past month, then that reading gets reset to zero when the meter is read at the end of the month. Did that peak occur during the night, thus extreme measures to save on outdoor lighting will pay handsomely? Or was it during the day? Nobody knows. Even the utility company doesn’t know. Shaving peak saves the energy customer large sums of money and saves the utility company the cost of larger infrastructure capacity, but nobody knows when the peak occurred for an individual customer. All the talk of approving or rejecting new sources of power and new transmission capacity and saving energy is done in ignorance of this.

And 99.9% of utility employees and energy geniuses don’t understand peak. They think that the “peak” use is the highest instantaneous use during any 15 minutes during the past year, which is nonsense – the highest instantaneous use during any 15 minutes or 17 minutes or 15 seconds during the past year are all equal, making the 15 minutes part of it all meaningless. The peak is actually the largest 15 minute advance in the meter reading that sums the energy used (20 minutes with some utility companies). With almost nobody understanding this, efforts to reduce peak use are often folly.

As to the question of elevators, electricity use (watts) is the product of volts x amps x power factor. A good voltmeter costs $20, a good ampmeter costs $50, but a device that measures power factor (the synchronicity of the peaks and valleys of volts and amps) and multiplies out to watts costs hundreds of dollars (except for some $11 ones for small plug-in loads), thus many people who say they measured power simply skip the power factor part, and get a meaningless result, including for research that you paid handsomely for.

Measurements are fairly easy for a steady load, such as a light that stays on. But if the electrical load is constantly changing, as with an elevator, the equipment needs to be able to record every second, or better every half second, or continuously. I could have bought a meter of the type utility companies use, but the resolution is larger than the small amount of electricity used by one ride on an elevator, and taking 10 or 100 rides to get a reliable meter reading isn’t practical. I could have tricked a utility meter into recording 10 or 100 times the reading (loop 10 or 100 turns of wire through the meter), but then I would have to run and read the meter before my test ride and then again before someone else decides to go for a ride – not practical.

I used a $5,000 power quality meter that records 30+ columns of data on what is going on in an electrical circuit, complete with timestamps of each reading (every half second), and used a watch to record when I started a test ride, so I could go find that ride in the data later, and also find and ignore rides taken by building occupants who didn’t happen to be going to the floor I wanted to go to, and who changed the weight in the cab by being there. Column E in the enclosed spreadsheet shows the “start” and “stop” times for various up and down rides, which I then summed to get the energy used for one ride. Watts gross are for all electricity used by the elevator, including controls and lights, and net is the energy used by the motor alone.

Only a few companies make these meters. After buying some of the less expensive ones and quickly returning them because of huge problems, I settled on Fluke brand: the leader in the industry. Turns out the instruction book had terrible errors in the diagrams that showed how to connect the meter to different types of electrical circuits. The tech people at the factory knew about the errors but never fixed them. Later, when the print book was replaced by a PDF on the factory’s website, the errors remained. I wondered how many other buyers knew enough to catch those errors – I imagine very few people buy these meters and still fewer use them and still fewer figured out the mistakes in the instructions and used the properly – almost nobody measures energy, despite it being somewhat important.

Despite being the industry leader, Fluke seems to have pioneered a terrible error in how electricity gets measured. Power factor used to be a number from zero to one, with one being perfect synchronicity and zero being opposite peaks and valleys. The peaks can “lead” or “lag” each other, and old meters used the words “lead” and “lag” to describe this. Fluke decided to simplify matters and describe lead (or lag?) as a negative power factor number, thus when that negative number is multiplied in the equipment by volts and amps (always positive numbers) the result is a negative number, implying the electrical load is supplying power to the grid, which of course is nonsense. When a positive number for one minute is added by the equipment to a negative number for some other minutes, the sum equals total garbage, yet this seems to be the best equipment available today. Since Fluke led the way with this mistake, other companies have followed. The last time I looked at the definition of “power factor” in Wikipedia it said power factor varies from zero to one, but then added on at the end was the claim about negative power factor meaning the load was sending electricity back to the generator, the exact same wording as on Fluke’s website.

Surely in trading there are examples of someone attempting to change reality, or change the definition of something in wide use to suit their purpose. Once upon a time I started keeping a list of all the different definitions of “inflation” by different experts, categorized by government or not, nobel prize winner or not, but soon gave up as there were too many definitions and too much vagueness to even make categories.

It is very strange that something as important as energy is measured every month by utility companies, yet the measurements are largely ignored by the industry in favor of estimates and models and hopeful claims. That is definitely the case for the heat pumps being installed now – most everyone claims they save energy, but when the energy used to make electricity for them is included, they use much, much more energy than the equipment they replace. But, with almost nobody measuring, and many measurements being difficult to get, things will probably continue along the same way for a long time.

No doubt there are many trading opportunities where few people measure because it is difficult to measure, or despite measurements being easy, nobody wants to see what the measurements show.

Humbert B. wonders:

How did you get building managers to give you access to the guts of so many elevators? Presumably it must have been part of a job. But what was the job? And why would so many buildings have given you access?

Henry Gifford responds:

The surface answer to how I got access to all those elevator equipment rooms is that once upon a time I owned apartment buildings in Manhattan and Brooklyn. Owners who knew that were immediately comfortable with me. When I got out of owning buildings (in NYC) because of corruption and rent control, I started a business fixing boilers to save energy and to run for years between breakdowns, thus I knew many building owners and managers.

The more fundamental answer is that it took two or three years (in my “spare” time) to learn to measure electricity used by an elevator, and figure out what equipment actually worked, and during that time I mentioned my project to people, most of who got fascinated with the idea of measuring something important which apparently had never been measured before, thus doors swung open, and one person told another, etc.

The wildest situation was a modern 50? story tall building near New York’s City Hall where the manager, who I never met in person, promised to tell the security guards at the door to let me in at 3AM, when few people would be using the elevators (residential building). I walked in and told the guy at the desk in the lobby why I was there. He reported that nobody had told him anything, but let me in anyhow, and unlocked the rooftop equipment rooms for me. This with all my equipment in a backpack where nobody could see what I was carrying. I basically talked my way in, like people in movies do. I already had data on lots of mid-rise buildings, and some super heavy duty warehouse elevators, but no fast elevators in very tall buildings, thus this rounded out my data set with data I felt lucky to get, as I never worked in buildings like that (corruption, unions, etc.).

Everyone who let me in took a risk that I might damage the equipment, leading to an expensive repair, as the service contracts state that nobody but the service provider touches anything or the owner writes a blank check.

Carder Dimitroff suggests:

See Mr. Gifford's wonderful book:

Buildings Don't Lie Better Buildings by Understanding Basic Building Science

Archives

Resources & Links

Search