Western Mountaineering vs. Ray Jardine: Sleeping Bag Temperature Ratings

[Ryan Jordan]

This was a fun analysis!

The dots represent every WM bag/quilt. Some are on top of each other, so a dot may represent more than one product.

The red line is Ray’s equation.

[Ryan Jordan]

(Maybe thru-hikers and mountaineers can agree to agree when nighttime temps are 45-55 deg F? 🤔) and then it all goes haywire?

[Jeff McWilliams]

Are WM’s bags EN or ISO rated?

I have a copy of Ray’s book somewhere on my bookshelf, but I think I’d trust WM’s temp ratings more than Ray’s.  They seem to have a stellar reputation.

[Brad Rogers]

Jeff – WM bags are EN rated.  The chart is in the FAQ section of the website.  It looks like overall, WM is a tad conservative compared to their EN Ratings.

https://www.westernmountaineering.com/faqs/

[Ben H.]

I assume the loft is as reported by WM. It would be interesting to see how the lines shift based on the the two EN ratings for each bag.

[Bob Shuff]

Ray was hardcore and thought a lot about picking the campsite. He also didn’t stay there long, which I’ve come to realize isn’t exactly my style. Maybe you just need to spiral more the Ray way to get by with less. Also maybe WM over builds their products, hence the stellar reputation.

[Scott Smith]

I think that we should replace the EN Standard with the WM Standard

[Brad Rogers]

Ha!  – Except we don’t know how WM calculates their ratings or how they compare between bags.   Some of their bags, like the Summerlite, come out exactly like the EN ratings.   Others are much more conservative than the EN ratings.   The EN rating system may not be perfect, but at least it’s consistent between bags and brands.

[Rex Sanders]

The EN rating system may not be perfect, but at least it’s consistent between bags and brands.

Uh, no.

See my interview with long-time gear creator Mike Glavin, under “Sleep System Standards Pitfalls”

I didn’t realize the full deficiencies of [the EN sleeping bag temperature rating standard] until I tried to use it for development and testing.

Mike went into more detail on his company’s web site:
https://zenbivy.com/blogs/news/why-our-sleeping-bags-are-no-longer-en-rated

Lab-certified EN test results are misleading because the test methods have very wide variances that make them unusable for both product development and comparing the thermal performance of one sleeping bag to another.
[Long explanation follows]

My advice: since makers and retailers almost never post loft measurements, use EN Comfort ratings (not the headline Limit ratings) to narrow your sleeping bag choices. Then look for trustworthy reviews (not found on Amazon or REI), ideally comparing the warmth of one bag to another. Or play the buy-try-return-repeat game.

I gave up on backpacking gear standards for many reasons.

— Rex

[Stephen Seeber]

I have been studying ISO 23537 for several months as I work to find a sleep system that keeps me warm in the winter.  I think when ISO 23537 is implemented correctly it provides a useful reference for comparing bags.

In the standard development, repeatability was evaluated by testing using 6 different thermal manikins and six different sleeping bags and then repeating each test three times.  They found the variation on each sleeping specimen to be 3.6%.  Close enough.

They then repeated this testing regimen at several other laboratories and found a 5% variation.

This seems like a pretty accurate test. Of course, the devil is in the details at the testing labs.

A thorough reading of the standard reveals that the authors are well aware of various limitations of the test method and provided appropriate warnings  This is to be expected because there are a variety of variables in the real world that cannot be incorporated into the laboratory test.  What can be controlled is the calibration process to conduct the test as well as the administration of the test itself.

The test calibration process covers the mannikin used for the test, the environment of the test chamber and critically the use of 5 reference sleeping bags.  It is a pretty rigorous calibration procedure.  Calibration is the responsibility of the testing lab. Not ISO or EN.  From the posting on the Zen Bivy website, I could not help but wonder if KSU was rigorously adhering to the calibration process and providing proper controls of the testing procedure.

This testing process is pretty complex.  The calibration process is demanding.  KSU charges $600 for testing a bag three times, as seen here. I would have expected such a test to cost a couple thousand dollars per bag. I don’t know how the published fees are sufficient for the level of investment and technician and administrative time required to test sleeping bags. Of course, KSU knows better than me what they need to run their lab.

The KSU documentation states that they calibrated using the reference sleeping bag set in 2019.  This statement suggests to me that this was the last time the reference calibration  was undertaken.  The standard does not say how often the reference sleeping bags must be tested to confirm system calibration.  The standard does say that measurement variance of the reference bags must be less than 5% of the published values.  It is unclear to me that a calibration frequency measured in years is appropriate for this level of accuracy.  The standard requires testing on a sleeping pad with an R-value of R4.8.  KSU states that they use a 1.5″ self-inflating pad.  You can see a picture of the pad in the link provided above.  I used such a pad years ago.  I don’t know what the R-value was, but I suspect it was well below what is required by the standard.  KSU states that they can rate bags for children.  ISO 23537 states that ratings for children cannot be accomplished.

In the Zen Bivy link that explains their abandonment of the standard, they seem to blame the standard.  However, I wonder if the KSU testing approach conformed to all the standard requirements?

I have gone through the experience of sending samples to a lab for testing and received back results that could not be correct.  I then questioned the lab until I understood how this could have happened.  Laboratory testing always requires a high level of dedication to instrumentation accuracy and conducting test procedures that meet all the requirements of the test standard.   If this does not happen, you can end up with test result deviations of the sort described by Mike Glavin.  The fact that the lab did not catch the variances that Mr. Glavin describes is cause for alarm.

I do agree that loft is the most critical piece of information about sleeping bag thermal performance.  Most sleeping bag manufacturers do not publish loft measurements.  Western Mountaineering does.  I wish more manufacturers would follow suit. However, each user will have individual needs that cannot be met simply by knowing the sleeping bag loft. This is a difficult problem to resolve except through experience.

I hope to have more to say about evaluating sleep system performance over the next few months.  I will say that my search for an appropriate winter sleeping system shows how difficult and expensive it can be to achieve comfort.  This applies to sleeping bags, sleeping pads and even tents.

For now, it’s back to REI to return a sleeping pad (it has gaps in its insulation that significantly degrade its performance). I had hoped to use the new pad in place of one with so much convection loss that it short circuits the closed cell pad I stuck below it to achieve a higher R-value.

 

[Robert Spencer]

I love charts and spreadsheets, but when it comes to sleeping bags, I guess I’m a buy and try guy as Rex says because I never pay attention to the EN ratings etc. I know there are many variables like cut, baffle construction, draft tubes, hoods, fill power, etc. but my go-to statistic is fill weight. From experience, 12 ounces of down is my minimum for summer Sierra travel, but 14 ounces is just right. I use this as a guiding principle and then make sure loft and temp rating are in the ballpark. Not saying this is the right way, it’s just what I do to land on a suitable bag for my needs.

[Rex Sanders]

In theory, ISO 23537 ratings might be a good way to compare sleeping bags.

In practice, so-called EN ratings have several problems, including:

– Inconsistent results from KSU, the “independent” test site set up by the US outdoor industry, and used by most US sleeping bag makers. A less scrupulous company could keep submitting nearly identical bags until they get close enough to the desired rating for marketing purposes.

– Very misleading marketing of the results.

– Claims that R-3 or even R-2 sleeping pads are good for “three-season” use when ISO 23537 requires R-4.8.

– Unrealistic test conditions, including no movement inside the bag, virtually no air currents, and 50% relative humidity.

The standard authors explained some testing limitations using technobabble buried inside a $120 document. But bag makers and retailers skip over all of that and market the WhizBang 20 sleeping bag as “good to 20 F” without qualifications, if the bag tests to 23 F under ISO 23537.

The standard got updated this year (ISO 23537-1:2022):

The main changes are as follows:
— update of Clause 3;
— update of the scope to exclude extreme climate conditions;
— revision of requirements for lower temperature limits;
— revision of test methods;
— revision of Clause 7;
— revision of the reference values of thermal resistance for calibration of thermal manikin.

Also:

In this document, consideration was given to the need to continue to reduce inter laboratory variability of the thermal testing and a number of test parameters have been tightened as a consequence.

I believe the applicable phrase is “caught red-handed.” Not clear how this revision might change future ISO 23537 ratings.

Bottom line: US consumers can’t rely on recent “EN ratings” for much of anything. We would be better off going back to loft measurements, which consumers could double-check at home with a ruler instead of a $100,000+ testing lab staffed by experts.

— Rex

[David Hartley]

I am disappointed in the large variability of the EN test results, but not surprised. I have bags that live up to (and some times exceed) their ratings and have had several that don’t. Certainly loft is eye-candy – nothing like a poofy down bag for the look of cozy warmth potential, but loft alone ignores the fit of the bag and other construction details. And of course the other variable – likely even greater variability than the EN testing issues – is the individual and if they are warm or cold sleepers. Another consideration is the length of the trip and the weather forecast (e.g. rain/humidity) – for me it seems like the coldest nights always follow several days of rain and clouds with no chance to air out my bag.

At the end of the day I have had to figure out what works for me from experience – I don’t look at loft per se – I look at ounces of >800 fill down and the fit of the bag. And the sleeping pad is also a huge deal for me (and another whole subject). My first down bag purchase years ago was a Marmot Pinnacle, with something like 22 oz of down (when Marmot started EN testing they reported a 22F comfort rating). But the bag was heavy (like 42 oz), had an overly generous cut, and at the time I was using a Z-lite foam pad – when temps dipped much below 30 degrees I was cold. But now, for spring/fall hikes I use a Mountain Equipment Co. Firelite with 17 oz of >800 fill down and a narrowish cut, and even though its EN comfort rating is 32F – I know from experience that I am good to the high twenties with a NeoAir All-Season pad – even after several days of crappy weather, and have been good into the upper teens by adding a 60g primaloft jacket when all the down clusters were at their poofiest. I also know that with 10-12 oz of >800 fill down I’m good to high 30s/40 degrees, with the fit of the bag much less important – but also that this warmth will be challenged by several days of rainy weather if I can’t air out the bag in the sun (I have started using a 30 degree synthetic quilt for this type of trip).

For someone just starting out I think the “Western Mountaineering” ratings are probably on the right track. If the budget allows – taking the time to discover and purchase from reputable/trusted companies (e.g. Western Mountaineering, Feather Friends). If the budget doesn’t permit that type of purchase (it certainly didn’t for me) then the EN ratings are still a place to start, while perhaps keeping an eye on how these trusted companies rate their bags vs the oz of down and the fit – understanding that you still have to learn what works for you through experience.

[Stephen Seeber]

Hi Rex:

I kind of agree with the use of loft.  Hard to check with a ruler, but easy to check with a device you can make at home in an hour and corresponds to ASTM F1932/F1932M.  It is also the best way to keep track of loft loss resulting from use.

I think the standard shows awareness of the potential for misuse of the standard and clearly state how it should be used. For example: Appendix E–Warning of misuse of temperature rating

The insulation of a sleeping bag varies widely with the conditions of use (wind, radiative ambience, posture and clothing of the sleeping bag user, ground insulation, eventual humidity in the sleeping bag etc.). The perception of cold of the user is also individually different (influence of acclimatization, physical and psychological state, food etc.).
The limiting temperatures of the range of utility as determined in this document only compare performance of sleeping bags with regard to standardized test conditions. They do not take into account all possible variations in conditions of use and in individual reactions, and therefore should be considered only as a guideline, that needs personal adaptation for practical use.
In particular, it shall be noted that the extreme temperature is a very theoretical limit. It shall therefore only be considered as a point of danger that should not be approached, unless the sleeping bag user has a wide personal experience. Furthermore, it has been determined through tests and practical experience that internal dimensions and the dimensions of the user have a significant effect on the performance of the sleeping bag.
The determination of the comfort temperature uses the available knowledge of published data and is based on the thermal balance of the whole body. The human body is very sensitive to local discomfort: a local thermal bridge might not influence the global insulation of the sleeping bag but might greatly affect the sensation of cold of the sleeping bag user. It shall be emphasized that the test method in this document does not provide any guarantee against local cooling.
The temperatures of the range of utility relate to indoor conditions. For outdoor use, wind can affect insulation of the bag to a large extent, especially if the shell fabric of the sleeping bag is air permeable.
In this document, sleeping bags are considered as dry. High moisture content can lower thermal performance.
© ISO

The industry does misuse the standards.  For example, they can and do advertise a bag to be good to its limit rating.  I think if a single number is to be used it should be the average of the limit and comfort rating. Or, publish the R-value resulting from the test. The average of comfort and limit values actually corresponds pretty well to Western Mountaineering ratings.

The standards are clearly not ignorant of the issues you raise but they cannot control the behavior of unscrupulous manufacturers.  I think it would be nice to have better insight into the KSU testing accuracy and process.

[GR]

Can you share more information/links about the device you can make to check this at home?

“I kind of agree with the use of loft.  Hard to check with a ruler, but easy to check with a device you can make at home in an hour and corresponds to ASTM F1932/F1932M.  It is also the best way to keep track of loft loss resulting from use.”

 

[bradmacmt]

I’m a die-hard WM fan, have used them going back 30 years with many models. Besides being conservative with claims, they are one of the few that uses loft measurements. But remember, loft isn’t just something you can measure with a ruler, it also has to do with the depth of the baffle construction – WM gets their loft numbers by adding together the physical baffle construction depths. That’s different than just sticking a ruler next to a bag laying flat. It really irks me a maker like FF does NOT offer loft measurements. Guess I’m just getting old…

[baja bob]

Under the spec chart EE provides a target loft for their down quilts.  I believe Katabatic does the same.

I’ve had WM bags and the loft measurement seems to be about what you get with a ruler.  What method do they use to calculate the loft?

[Stephen Seeber]

How to measure sleeping bag loft. ASTM F1932/F1932M Standard Test Method for Measuring Sleeping Bag Loft.

When you measure fabric thickness, you must apply a bit of weight to eliminate random changes in puffiness.  In this standard, you need a rigid structure to support a ruler.  The ruler rests on a 12″ diameter disk.  The total weight of the ruler plus the disk must be 50 grams. The ruler must have 1/16-inch graduations.  The ruler must be long enough to measure from the bottom (or top) of the horizontal structure (just be consistent) to the top of the surface on which the sleeping bag rests.  Before doing the test, you want to shake up the bag and then let it sit in a horizontal position for 24 hours.  Then, complete the measurement.  I measure the widest part of the mummy bag, the middle of the bag, and just before the foot box and average the three results.  The photograph below shows the measurement setup. Here, I measured a WM Antelope bag. The loft is supposed to be 7″.  I measured 6.8″.  My WM Alpinelite is supposed to be 5.5″.  I measured 5.1″ after washing and drying.  I wonder if they owe me another .4″?  I don’t know what their manufacturing tolerance is.

The disk is a .676 expanded polystyrene foam board.  The trick is to come up with a combination of ruler and disk that weighs exactly 50 grams.  I got it right on the 3rd disk material. The ruler I used is from the C-Thru Ruler Company, 18″.

 

[Bonzo]

(Maybe thru-hikers and mountaineers can agree to agree when nighttime temps are 45-55 deg F? 🤔) and then it all goes haywire?

Gott im Himmel, I absolutely do not want to be on a hill when it’s 55°.  A few weeks ago I got stuck at 6500′ with a 0° bag and it was 42° ambient overnight, and I thought I was going to die from sweating so much…or from turning into a glacier snack. 🤣

But in all seriousness: so much of this is subjective.  One of my hiking partners is so cold-sensitive that the aforementioned 0° seems cold to her at any sub-freezing temp; another one that’s the exact same size and shape is firmly convinced that there’s no location on earth in which that bag is useful, because it’s too warm.  I waffle back and forth…so my solution has been very simple: own lots of sleeping bags and pick whichever seems the coziest at the moment.

[Ben H.]

Complaints about variability in en ratings is laughable when the alternative isn’t a better rating, it’s pure marketing whimsy. Sure an unethical manufacturer could keep sending a bag in for measurement and choose the best rating …or… They could do what they are currently doing and just make up a rating for the bag. Guess which one an unethical manufacturer will choose. Like all standards, I take complaints from manufacturers as frustration over limiting ability to manipulate numbers. Loft may be a better measurement but it doesn’t mean much to the average consumer. I actually wish they would create a similar standard for mats. You see a lot more novices with undersized mats than understand bags.

[YoPrawn]

What gives with the Summerlite sleeping bag? Why did WM deviate from their “conservative” rating system on that particular item? Sure, some warm sleepers or people wearing clothes and in a double skin shelter, could feel good in that thing at 32* F, but that would be the exception and not a standard comfort rating.

It’s a great bag, and warm enough if you know the limits of it, but 32* F comfort rating? No way. Maybe 40* F would be closer to accurate.

For now on, I only go by the fill rating and weight, along with size of quilt/bag. Baffled and differential cuts are a given for high-end stuff, so the down fill and size seem to be the easiest to factor.

My Katabatic Gear Flex 22 has darn near close to double the down fill as the WM Summerlite, rated at 22 degrees, and it is a quilt with no hood. No way the sleeping bag with far less down fill will come that close in temp rating. It’s a true 22* quilt for comfort.

[Scott Smith]

Back to one of the guiding principles…everyone needs to understand that everyone sleeps different..

I’ve been in a Summerlite, at 22 degrees and slept warm and cozy. EXPED 4.9 R value rated mat…and my Hornet Elite 2p…I have a Versalite, that I took to Bhutan, however, haven’t used it since..

[DriCamp]

Question. How to apply the WM regression line to a quilt? I own an EE RevX and it lofts to about 3″ (I measured 6″ of loft when the quilt is doubled, and lying between two sheets of foamcore). Certainly the quilt can’t do 15F as per the regression line for 3″. My impression is that the quilt peters out around or before freezing (atop a NeoAir). Is a “quilt compensation” required to apply the formula?

[Stephen Seeber]

Hi Dricamp:  Based on data from the EE website, ratings are based on a single quilt thickness.  WM is based on a complete sleeping bag with a person in it.  I don’t know what either company assumes about pad R-value.  You cannot use Ryan’s regression for your EE quilt.  Is your quilt Apex or down?  EE Apex quilt ratings range from 50F to 20F for Apex.  This corresponds to Apex weights of 2 osy to 8 osy in 2 osy increments. Each additional 2 osy produces a ratings drop of 10F.  EE Apex down quilt ratings range from 50F to -10F.  .5″ of increased down loft drop the quilt rating by 10F.  I have no idea where EE gets this information or how reliable it might be.  WM ratings generally correspond pretty well with ISO /EN ratings that they publish on their website.  There are some outliers such as the Summerlite and the various WM bags containing 7″ loft.  For complete sleeping bags with a person and pad of unknown characteristics, each additional 1.09″ of insulation drops the WM rating by 10F.  Using ISO/EN rating data and WM loft data, each additional 1.28″ of down drops the rating by 10F.

[Ben H.]

I’m not sure where you are getting 1.09″ of loft per 10°F change in rating. According to Ryan’s chart above WM changes their rating at a rate of 20.1 °F per inch of loft which is almost exactly inline with EE. Jardine uses a more aggressive 40°F per inch of loft.

[Author]

Posts