## The Downside of Adopting the Metric System

Posted on December 9, 2014  Comments (4)

The only downside of adopting the metric system is less control over room temperature (based on my experience). Every ºC = ºF * 1.8 so have less control (when using only integers to control temperature as is the case in my experience).

Granted this could be solved easily by using .5 (option in air conditioning and heating controllers but in my experience they don’t) for Celsius. For Fahrenheit this works out to enough control for me. For Celsius in a fair number (lets say 15%) of systems it is a bit uncomfortable.

The specific circumstances add greatly to creating a problem. My guess is those that annoy me swing even further than 1 ºC, they move further in one direction in order to not turn on and off all the time. So maybe that moves to swings of 2 or 3 ºC at the measurement point. But that is another issue, the measurement on home (or hotel) systems is often 1 reader so the variation is often greater in other locations.

Add to that the imprecision of their measures, I don’t have good data, but I am confident that the measurement error is fairly high. I am pretty comfortable at about 25ºC for air conditioning. But in some places I am cold at 27º and others I am warm at 23º. It could be me, but I don’t think so (most of the time – sometimes it is me).

A long time ago I had some imprecise portable temperature gauge and while I wouldn’t stake my life on results based on it, it confirmed my feelings (when I felt it was warmer than the local reading said my device agreed and when I felt it was colder my device agreed). Hardly scientifically valid proof, but it made me more comfortable trusting my opinion on this matter anyway.

My guess is in a unit using ºF you often can be 4 or 5 degrees off (or more) in different locations. For some people that might be ok. But for me that often starts to be uncomfortable. If you convert the issue to that time 1.8 it is noticably worse.

Now in reality I don’t think it expands quite that much. While the manufactures balance the confusion of adding .5 to a Celsius controller and decide not to, I would think they don’t swing 1.8 times as far (in heating or cooling in order to not turn on and off all the time), but it is still let precise than using Fahrenheit integers. I believe (hope) they set their internal dynamics not based only on integers but could say for example turn off .5º past the setting and turn on when the conditions are .5º worse than the setting (so .5º too warm in the case of air conditioning, for example).

It is still lame the USA fails to adopt the metric system, but reducing this problem in the USA is one small benefit of holding off 🙂 I wonder if 1 in a million, 1 in 10 million… up to 1 in 7.2 billion people (just me, all alone in the world) have my concern for the lack of precision of heating and air conditioners when using the metric system.

4 Responses to “The Downside of Adopting the Metric System”

1. Anonymous
December 27th, 2014 @ 8:33 pm

That 1.8 degrees of control is a valid concern, especially on things like process controllers that only display 4 characters (ex: ‘999.9’). If you set the unit to read in metric then you lose some of your resolution.
(Now whether that additional digit is false accuracy or not will depend on you’re equipment. Just because the reading is steady doesn’t mean that it is accurate).
For a metric-only process controller to have the same precision, its display would have to have an additional character, the .05 as you said.

As for your personal comfort level varying randomly with the temperature, that could be due to humidity or your activity level. Either that or you’re just weird.

2. John Moss
January 6th, 2015 @ 8:40 am

Could this problem be traced back to software in more complex systems? The algorithms used in Celsius and Fahrenheit controllers are probably the same.
The normal programmers device used to prevent controls hunting might well be a simple “switch” that operates if there is a deviation of the measured value from the controlled value of a certain number of units. Integers have advantages in simple programs. If the algorithm was written assuming a design for a Fahrenheit based system then using the same algorithm without adaption leads to an increased control deviation of 1.8 in Celsius times the deviation in Fahrenheit.
I stand to be corrected of course, but the early scientists who decided on the temperature scales favoured an interval that was just noticeable or not quite noticeable, just as they did when defining the Decibel for sound levels.

3. Andy
January 26th, 2015 @ 5:31 am

John, I think one of the reasons that the UK has not gone metric is that the US still uses the old Imperial system. So if you can get the US to change to metric then I’ll lobby the manufactures to add a 0.5 degree control to their devices.

The human perception of temperature is very subjective. There’s an experiment that you can perform with 3 containers. One warm, one cold and one average. You put one hand in the warm and one hand in the cold and leave them a few seconds. The put both hands into the average temperature container. You hands will perceive the temperatures differently even though they are in the same container.

4. Robin
February 6th, 2015 @ 4:49 am

I vouch for a common system, which stands to be the SI one but as you’ve mentioned Imperial system has better resolution and I agree to that in most contexts like temperature and length (inches, miles,etc.) But I believe that certain implementation of units become a convention like usage of degree Fahrenheit in medical purposes.

Hey, I have used metric system all my life though being an avid physics student and engineer, I have come across Imperial system too, mostly while solving problems from American Authors.