With the news here in Europe full of the effect of the war in Ukraine on gas supplies and consequently, prices, there it was on the radio news: a unit of measurement so uniquely British that nobody uses it in the real world and nobody even has a clue what it really means. We’re speaking of the Therm, one of those words from our grandparents’ era of coal gas powered Belling cookers and Geyser water heaters hanging over the bath, which has somehow hung on in the popular imagination as a mysterious unit of domestic gas referred to only in the mass market news media. What on earth is a therm, and why are we still hearing it on the news in the UK?
You can’t Buy A Therm
Asking the internet what a therm is reveals the answer, it’s 100,000 BTU. What’s a BTU? A British Thermal Unit, another anachronistic measurement five decades after the UK went metric, it’s the amount of energy required to raise a pound of water by a degree Fahrenheit. Which in turn is about 1,054 joules, in today’s measurements. So a therm is thus a unit of energy, can we take a look at our gas meters and see how many of them we’ve used this winter? Not so fast, because gas isn’t sold by the therm. Older gas meters had cubic feet on them, and we’re guessing that now they’re calibrated in cubic meters. We can’t even buy a therm of gas, so why on earth are the British media still using it?
To answer that question it’s fair to say that there are two reasons for the warm and cosy grip of the therm on the national discourse here. The first is that surprisingly, wholesale gas is traded in therms, so while we consumers buy it by volume, our utility companies buy it by energy. At the time of writing a therm of wholesale natural gas costs about £2.60 (about $3.42) to them, but given the geopolitical situation of the moment it’s anybody’s guess where it’ll be tomorrow. So when the price of gas is quoted in therms on the news it’s because somewhere a utility company is still buying the things — who knew!
Mr. Therm, He’s Hot Stuff!
But the other reason for the news media’s fondness for the term is cultural. We’d never have heard of the therm and it would have languished as an obscure engineer’s unit of energy derived from gas, were it not for the gas industry’s mascot. Mr. Therm was a cartoon character used to promote the gas industry and gas products from the 1930s until the 1970s, and his ubiquity gave the word a hold over the popular imagination that must still be there for the older generations. Perhaps the papers and newsrooms still fondly cling to Mr. Therm, or more likely, they recognise that it’s mostly older people who still buy printed newspapers.
The therm then, a unit of measurement nobody uses and nobody knows what it is, but one which lingers on in a corner of the gas industry and in fond memories of a world long past. Perhaps it’s best to think of it as a customary measurement in an old country, like a pint of beer — the real unit is the familiar sized glass. We’ll keep our 29.3 kWh, thank you very much.
Header: George Shuklin, Public domain.
I wonder how many therms it takes to raise a cubic smoot of water one degree Rankine?
Oh, it’s about 23.1407, because smoots represent logarithmic volumes to base e and a Rankine is proportional to pi. Hence it requires e^(pi) therms. I thought everyone knew this ;-) .
And here I thought a Smoot was a unit of length (5’7″) used to measure the Harvard Bridge.
+1
That’s Smoot the Elder.
One of the most useful classes I took in college was Thermodynamics. I’ve only ever needed to use PV=NrT once in my career, but the teacher made a point of giving examples and test questions in odd units so we would get comfortable with unit conversion. I have fond memories of converting furlongs per fortnight and stone-cubits to m/s and joules.
A class I took 45 years ago. On final ” how much is a ton of refrigerant.” I put 144,000 BTU. Nope it’s 2000 pounds.
Turns out: about 1/9 of a therm.
Checks out. Looks like 174 cubic feet in a cubic smoot (smoot being 5’7″). Water is about 62 pounds per cubic foot, so 10,791 pounds of water in a cubic smoot. That’s about 0.107 therms to raise it 1 degree.
I think your problem is that a therm is a rate, like km/hr. You want to ask how long it will take at a given rate, or what rate you require for a given timeframe.
therm isn’t a rate, it’s a unit of energy. about 105.5MJ in SI, or 29.3kWh in convenient
Actually, in PL (and most probably in the rest of EU) consumers buy gas by watt-hours. Yes, meters measure volume, and there are complex regulations in place, but we have kWh on our bills.
yep, can confirm that is the case in Germany, as well, and I imagine that it’s an EN standard that mandates that.
I wonder how it is being measured, seeing that gas from the line is not entirely consistent in energy per mass or volume. There’s different gases in there, coming in at different pressures and temperatures.
My bill in the US is measured in volume with a variable “heat factor” for energy content. The billing is done in therms.
It is measured by volume. Pressure variations are corrected by pressure regulator just before the meter, so the pressure variations are minimal. Temperature variations are present, bu I guess that they are just ignored.
Natural gas composition is regulated, but not strictly controlled. If there’s too much of something, they mix in gas from another source and dilute it, but with biogas and hydrogen production, they’re also relaxing the regulations and allowing more CO2 and Hydrogen in. There’s anything from C1 to C5 molecules in there, leading to a lower “methane number” which is sometimes a problem for generators because they start knocking/detonating on low quality natural gas. Then there’s also N2 and sometimes couple percentage points of Helium in it – but that’s usually separated.
In any case, the energy contents of the gas is anything but constant.
In the UK well before ubiquitous natural gas, we had “town gas” made in the local gasworks.
This was made from coal.
It contained some methane, hydrogen, probably a bit of SO2, CO and whatnot.
We were billed by therms, and I suppose someone would measure the gasworks output to find out what the therm/cu ft relationship was, from time to time.
The kWh/m^3 value is part of the gas bill. Can be different from region to region. Tolerances are specified in national norms.
Same in HR. Gas is measured by volume (cubic meters), but price is for energy unit, and on the bill volume is converted in energy using fixed coefficient (cca 9.74 kWh/m3).
The UK also bills in kWh (and has done for many years). Metering is in cubic metres or cubic feet.
My meter measures in cubic feet so the calculation to produce kWh is (acccording to my bill):
Amount used in cubic feet × 2.83 × calorific value (varies from bill to bill, but is around 40) × volume correction (1.02264) ÷ 3.6
I just remembered, when you go to an LNG/CNG filling station to fuel up a car, they charge you per kilogram. A kilo of gas is roughly equivalent in use to a liter of gasoline.
Where “roughly equivalent” means “factor of about 1.6”
1 kg CNG ~ 54 MJ (depending on methane fraction). 1 liter of gasoline ~34 MJ (depending on ethanol adulterant)
I find the heating value of gasoline quoted at 36.4 – 49.6 MJ/kg.
Oh, sorry, liter versus kilogram.
So, actually it’s easier to think about a Therm and a BTU in metric rather than Steam-punk units (though Steam-punk is fun). 1054J = 1 BTU approx = 1KJ. So, a Therm is 0.1GJ. Rats, I’ve reduced the article to the size of a tweet ;-) .
It would be just one more conversion constant to metric in my spreadsheets if I ever interacted with it. Mixing units is one of the lamest parts of engineering.Too much time in too many careers has been spent learning two sets of measurements until intuitive and grinding rough conversions into professionals’ bones. It can feel good to instantaneously grasp a thou measurement in microns (among dozens of other relationships), and understand what they both mean, but why?
In the UK, there’s a slow, but steady progression towards metric and even though I was only taught metric when at school in the 1970s and 1980s; most people in the UK are fairly inconsistent. eg
Petrol/ Diesel in Litres.
UHT Milk, Carbonated fizzy drinks, Fruit Juice, beer in tins, medicine, perfume and creams; methylated Spirits: all Litres or milliLitres.
Pasteurised milk, bottled milk, beer in pubs: pints.
Cooking generally done in ml, grammes; though older generations will do it in ounces / fluid ounces.
Height: Feet and inches (but I do cm).
Wood-working: feet and inches.
Other lengths: metric.
Weight: Stones and Pounds (14 pounds = 1 Stone) (but I do kg).
Energy: kWh normally, Power: W. Torque: usually Foot-pounds, but Nm is taking over.
Distance: metres, up to about 1km, then miles, unless running or cycling (e.g. a 10K).
Speed limits: miles per hour, but cycling often km/h.
The interesting part is that if you buy a ruler or measuring tape in the UK, it will have cm down one side, and inches down the other. Even an average person is quite capable of thinking in both.
Personally I use which ever unit is most convenient at the time, unless I have to do any calculations, in which case I convert everything to SI units. I can only do maths in base 10.
>I can only do maths in base 10.
I find that to be the difference with people who did not learn their fractions properly in primary school. It’s a pity, since the same skills and methods are required later for solving math problems analytically, and then engineering problems, unit conversions…
I found that the 16-times table I learned for lbs and ozs, invaluable when I became a software person and had to work with hex values. Still do, 50 years on.
“Mixing units is one of the lamest parts of engineering.”
Mixing units is one of the *most important* parts of engineering. Without the unit, it’s just math.
“It can feel good to instantaneously grasp a thou measurement in microns (among dozens of other relationships), and understand what they both mean, but why?”
Metric mixes units just as much as anything else. Just because “it’s all powers of 10” doesn’t make it any easier to have to remember typically at *least* 6-7 different “wacko greek/latin to number” mappings. And when you combine different derived units, you just end up with the same problem, but now you need to memorize *names* rather than units: thermal conductivity, R-value, thermal resistance. You can’t give me just a number there, you’ve got to give the units and then I need to go remember what the heck it actually is.
(R-value’s a really good example, actually, because everyone just effing lists the freaking *number*, which is a terrible thing to do for *any* unit-ful measure.)
Heck, even if you don’t give a convenience unit a name, you still use them. You know roughly what room temperature is. You know roughly the size of a kilowatt-hour. You know how big a 19″ rack is. You know roughly how big a building story is, or a city block. You know roughly how big an oil barrel is (to within a factor of 2, at least).
Mixing units *in a design* is insane. That’s nuts. But being able to connect convenience units to design units intuitively and quickly is really just basic estimation. And estimation is really just connecting numbers to physical objects. Which is engineering.
Giving numbers without units is the practice because half the population can’t understand the difference between watt-hours and watts per hour, and will sometimes argue back if you attempt to correct them.
It would just confuse them. Instead you give them a named number like “R value”, a formula for using it, preferably in the form of a nice mnemonic phrase, and repeat it until they can repeat it back to you without error. That’s the basics of vocational training.
That’s the entire reason for reference units and relative measures. You quote noise figures in dB, relative to a reference value. The dumb part is quoting R values, with no units, and not everyone agreeing to what they’re in reference to.
I’ve found that it’s relatively easy to explain the difference between WH and Watts per Hour if you write them out mathematically and show the math to people…
A Watt Hour is a measure of Watts sustained for a number of Hours, thus W x H.
Watts per hour is a measure of how many Watts are delivered per hour, thus W / H.
Most people are perfectly capable of understanding the difference between 3 x 4 and 3 / 4 as they result in 12 and 0.75, respectively and I’ve not encountered too many people who have completed the fourth grade and still have difficulty distinguishing those two operations.
“We’ll keep our 29.3 kWh, thank you very much.”
The irony involved in saying you prefer a *different* non-SI unit over some other non-SI unit…
(kWh is not an SI unit because hour is not an SI unit – it’s “approved for use” with SI units but it’s not an SI unit. Given SI’s bizarro trend toward cutting down on “approved units” one imagines it may only be a matter of time)
In official reporting and statistics, Megajoules are typical. Of course nobody can relate a Megajoule to anything.
I can. As with all units, it’s only a matter of how many conparisons you have in your head.
The Watt should be retired completely. It’s a joke that Kilowatthour, something that means “Joule per second multiplied by seconds” and is completely redundant, is an extremely common unit. There’s literally no other common unit like this.
You can also just literally replace “W” with “J/s” on the packaging of all electric devices and it’s done. It would be extremely easy to fix this.
The problem is that a Joule is an inconveniently small unit, and it makes electrical engineering difficult. For example, if you know that your cellphone battery has a nominal voltage of 3.7 Volts and a capacity of 1 Ah, you can’t just multiply them together to get 3.7 Watt-hours, no, you have to multiply by 3,600 to get 13,320 Joules, or 13.32 kJ. Not only is this cumbersome, it is also pointless because you then have to work it backwards to do anything with the number.
We’re not really interested in what happens over a time scale of seconds, so we’d rather not use units which are dimensioned in seconds, just like we don’t want to buy milk in cubic meters, or plan our road trips in centimeters.
Example #2: an electric car consumes 1,224 Megajoules per mile. If you drive for 10 miles, how many hours does it take to recharge with a charger that delivers 1 kJ per second?
Compare: an electric car consumes .34 kWh per mile. If you drive for 10 miles, how many hours does it take to recharge with a charger that delivers 1 kW?
Oh, sorry, I meant to type 1.224 MJ not 1,224 MJ. A factor of a 1,000 error – whoops!
It happens to the best of us – which just goes to show why it is a bad idea to deal with units that force you to go back and forth multiple orders of magnitude to calculate stuff.
This happens a lot in mechanical/structural engineering where you have to deal with millimeters, meters, their counterparts in area and volume, and then mega and gigapascals, so there’s all different powers of ten all over the place. One unit is measuring in increments of the weight of a feather spread over a coffee table, while the other units are either the size of a small grain of rice – or the size of a washing machine. Errors here will cause bridges to collapse and airplanes to fall out of the sky.
Wh is far less non-SI than therm. It has been introduced (at least at customer level) to gas trading to make gas heating comparable to electric. Interestingly electricity hasn’t moved to joules.
“Wh is far less non-SI than therm.”
For *now*. SI’s been killing off completely reasonable convenience units left and right, the bastards. First, they’ve come for your nautical miles, knots, bars and barns, next they’ll be comin’ for your hectares, days, and years!
I’m planning my 2022 Hecto-Day party this Sunday! Guests should arrive at 6 deci-days o’clock.
I cried when they took away the myria (10^4).
Hell *micron* is a non-SI unit! They revoked it back in 1967, and absolutely no one gives a damn.
It’s each to their own. One country sells drinks in centiliters, another in milliliters. One sells televisions by the centimeter, another in inches.
Technically speaking, every country is SI metric – they each just have their own choice of convenience units. Even the US.
I think you may have missed the element of humour in this line.
I was actually applauding it (and then completely killing the joke by over-explaining it).
haha matter of time
FYI, in the US, residential gas is still sold by the Therm even though the gas meter measures in cubic feet.
Not where I live in the Midwest. They announced it was to “simplify” billing, which means that they will use less energy dense gas as it will cost less. Same reason they don’t sell ethanol contaminated gasoline by the energy it contains (far less) and keep selling it by the gallon. I know the suppliers are getting to lie because the “smart meter” they put on the house results in bills estimating how much electricity my gas-fueled water heater uses. One lie begets another.
My (US) gas bill reads in “hundred cubic feet” (actually down to a fraction of a cubic foot on the actual meter) and the billing is done in therms. There is a “heat factor” multiplier on the bill for the energy content during each period. That multiplier is always somewhere around 1.0 (1.047 this month) so 100 cubic feet of natural gas = roughly 1 therm = 100,000 BTU = roughly 1,000 joules.
There has to be a mistake. If a therm raises 1 gallon of water by 1 degree Fahrenheit, and it cost ~3 per therm, it would cost a fortune to cook dinner, or even make a pot of tea. To raise a quart of water from room temperature to boiling would cost £100! Perhaps that’s supposed to be £2.60 per kilo-therm or mega-therm?
If it measures in feet instead of meters, wouldn’t it be a “feeter” instead of a meter? :P
“At the time of writing a therm of wholesale natural gas costs about £260 (about $342) to them”.
Not really, you mixed up pence and pounds…
btw: natural gas is usually traded in MMBTU units. 1 MMBTU = 1 decatherm = 1,000,000 Btu ≈ 1 GJ.
I thought something seemed off.
My gas bill in upstate NY reports the number therms used.
I believe my downstate NY bill also uses therms.
Some months I use 0 therms, and others I use 1 therm. So it’s kind of like billing the correct amount using PWM.
If you go to the source document linked, it’s pence per therm, not pounds. In upstate NY last March, a therm cost me about $0.50 to purchase and $0.27 for delivery fees. About 58 pence.
You’re not wrong. Thanks.
Of course the real question on everyone’s mind is just how to deliver enough Therms to the Kremlin so as to make a beautiful lake of molten glass without said energy being returned to sender.
It’s even more complicated than that: the calorific value of natural gas varies, so it isn’t a constant multiplier between cubic metres (or cubic feet for old meters) and Therms
See “Step 3” here: https://www.gov.uk/guidance/gas-meter-readings-and-bill-calculation#calculation-of-gas-bill
In a lot of ways it makes sense that you’re buying in units of energy, although I imagine it makes metering complicated.
In California, my PG&E Gas Bill is actually expressed in Therms. They even show the conversion from ft^3 to therms on the bill because the meter is still calibrated in (you guessed it) ft^3.
I’m not defending this practice, just saying that you can, in fact (at least in California) buy a therm of gas and I buy several of them each month.
I suggest two chickens and one sheep. You can also take one cat from this dirty herd.
Still therms here on the bill for natural gas in US. So nobody uses it any more? … Well not true :) . Still used. BTUs and therms are still around. Not many here would know what a Joule is either.
It’s a thing you can quite easily bring up in normal conversation though, like “My watch has 300 Joules”
Unless you’re talking about the big hunk of crap with the indiglo that needs the CR2032 then it’s ~2000 Joules.
The way I relate Joules is: 100 Joules kills a man, 10 Joules kills a rat, 1 Joule makes you curse a lot.
Well yeah, but yanks still use miles and pounds and hands to measure things, which, in a world where everyone else saw the light and moved on, is completely bonkers.
I would not classify the BTU as an “anachronistic measurement” in the U.S. with respect to HVAC appliances. Go shopping for an air conditioning unit and you will find them rated in BTU’s or tons. A heating furnace or boiler is rated in BTU’s as well.
Well, pretty much all units used in the US are anachronistic.
The units used in the US are fitted to the purpose. No one needs a simple conversion from inches to miles; there is no need to require scientific notation or a more complex series of suffixes to hide the scientific notation for what is clearly handled in the name of the unit. What is useful in the “anachronistic” series is not missing a suffix change and being off by three orders of magnitude. This is why they have lasted so long.
Also useful is that they are typically divisible by factors of 2 – 2 cups in a pint, 2 pints in a quart (that “qua” tells the next step) four quarts in a gallon. Each one scaled to the typical application for it. No one cares how many cups are in a cubic foot or a cubic mile. But they do know the magnitude with a sufficient accuracy using only one or two digits. Anyway, it was easy enough to learn in 3rd grade, so if you are as smart as a 3rd grade child you can manage it as well.
It’s important to note one of the main places where you get differences between US and metric is actually really philosophical, not “imperial vs metric.” US standards groups (e.g. SAE) preferred labels rather than true dimensions. As in, you get “#22 AWG” wire or a “#4 screw” rather than literally specifying the dimensions, and the scaling there tends to be ratiometric rather than linear steps (hence the reason for labels).
Even if the US standards groups had, for instance, chosen a metric basis for 36 AWG wire, you’d still end up with a “AWG to ISO” table.
And the next level of ridicule-US is that the actual unit is BTUs per hour and tons(of ice) per day
Here in Western Canada, Natural gas is billed in GJ (GigaJoules), so a lot of billing unit diversity…
Canada is all aboot diversity. In Ontario, Enbridge just bills by the cubic metre (as indicated on the gas meter, eh).
No barometric pressure or altitude adjustment. No adjustment for heating value. Just the straight meter reading.
Oh, and Canada is all about taxes and fees too: the cost of the gas is less than half the billed amount.
So one Therm is one dMBTU or one khBTU, easy. Has nothing to do with kings feet and how far apart carriage horses were in Rome.
Now … What’s a BTU again???
1 BTU is about one kilo-Joule ;)
What is joules per second?
I’d ask Thermo-Man what a therm is. I have a feeling he is both an authority about and knows the answer. :-)
If we’re doing arcane energy units, don’t forget the Quad = 10^15 BTU. Commonly used in totaling up a country’s, or the world’s, energy use or production. But for this use, exajoule ( 10^18 joules ) has the advantage of sounding cooler.
It’s to ensure customers can’t figure out the basis of their supplier’s charging regime.
The therm describes the energy purchased. 100,000 Btu or 1.055 × 10^8 joules and is roughly the energy in 1 cubic foot of natural gas at standard temperature and pressure. I love scientific notation on my fuel bill. Not.
A typical residential furnace is between 40,000BTU/hour and 200,000 BTU/hour so the “therm” consumption is between 0.4 and 2 therms per hour making single digit precision and using no scientific notation sufficient for home heating calculations.
The amount of bullshit contained in an electron in milligrams per second. Only the second scummiest profession on the planet right now.
Reminds me of the good old times in high school when the physics teacher would forget to specify which unit the answer should be in so we’d denote velocity in furlongs per fortnight.