What’s the benefit of synthetic?

Brett – Synthetic insulations comes from the factory with a fixed density to optimize its conductive, convective, and radiative heat losses. Down has variable density depending on its fill power and its density in the baffles.

Natural convection heat loss is negligible even at the lowest down density (highest loft). The conductive heat loss is dependent on the thickness of the insulation and so the thicker the insulation the less the heat loss. Radiative heat loss is decreases with higher bulk density up to a point. The optimal design of down baffles is a sweet spot just before the point where further increases in density, to offset radiative heat loss, occur at the expense of increased conductive heat loss.

Ron – A 20% overfill should yield the maximum thermal efficiency potential of down. According to my calculations 800 fill goose down, at this density, will be warmer than any synthetic insulation on the market. I am at a loss to explain why Ayce said that a down bag would have to be thicker to achieve the same warmth as an average synthetic.

Harold – Set the thermostat in your house to 70F. Set a heating pad to its highest setting. Use your own jackets or borrow some that are close to what Acye tested. Put the jacket over the heating pad. Put a thermometer on top of the jacket being tested. Wait a few hours for thermal equilibrium to set in and then record the thermometer reading. Do this for each jacket you have. The jacket with the lowest temperature on top will have the highest insulation. This simple home test won't tell you the insulation value, but it will give you an unbiased independent assessment of the relative warmth for each garment.

5/25/07 Simple Home Test Experiment:

New Balance Fugu (800 Down 1.5")
Cableas Jacket (650 Down 1.5")
Wild Things Jacket/Vest-Red (Primaloft1 1.2")
MEC Magma Jacket (Primaloft1 1.2")
Wild Things Primaloft Jacket-Red (Primaloft1 .6")
Wild Things Primaloft Vest-Red (Primaloft1 .6")
MEC Magma Pants (Primaloft1 .6")
Patagonia Micropuff Vest (Polargaurd Delta .6")
Patagonia Micropuff Pullover (Polargaurd Delta .6")
Patagonia Micropuff Jacket (Polargaurd Delta .6")
Wild Things Primaloft Jacket-Yellow (Primaloft1 .6")
Wild Things Primaloft Vest-Yellow (Primaloft1 .6")
MH Polartec 300 (.250")
Patagonia Black Puffball Vest (Thermolite Micro .156")

% of ideal is the ratio of the ambient room temperature / jacket top temperature

The Fugu down jacket is 800 fill and the Cabelas down jacket is 650 fill. They are sewn through construction. 1 1/2" is the maximum loft. The baffles average 5" wide. Only approximately 2" of the 5" is at the maximum loft. 1 1/2" on either side it ramps up from a few mm to 1 1/2". My crude estimate is that the average loft is about 70% of the max or 1.05". This compares to the Primaloft One MEC and the Primaloft One New Wild Things vest and jacket combination. Considering that the effective lofts are comparable, both down garments displayed significantly higher thermal resistance than the best synthetic on the market.

The yellow Wild Things Primaloft jacket / vest had been used on two multi-month expeditions and had been stuffed daily. The red Wild Things Primaloft jacket / vest is new.

The most striking observation to me is not so much the graph, but how much loss of insulative value the primaloft jacket suffered.

Richard,

I'm glad you did that experiment. An experiment is worth many many calculations.

With a little additional information, your home test should be quite convincing for ballpark numbers — certainly enough to resolve between the factor of 4 or 5 differences between claims. Do you know the approximate temperature of the hot pad? If the hot pad control setting fixes the power output and the geometry in each case was roughly the same (or, if you lay the garments flat on the pad, and you assume the power output of the pad is evenly distributed over area), you can obtain relative R values for each garment by calculating the temperature gradient.

Is it correct to read from your graph that the temperature on top of the fugu piece was 70 deg/.93 = 75 deg F, and on top of the WT vest or jacket 70/.81=86 deg F? Or did you use deg C, or K for temperature units?

If the heat pad temperature is around 110 F, you have temperature gradients of 35 degrees and 24 degrees, respectively, for down and .6" PL — so the 1.5" down is about 1.5 times as insulating as the .6" PL (2.2 clo). That is about 3.3 clo, closer to Ayce's experience. If the hot pad temperature is lower, say 90 deg F, then you have temperature gradients of 15 and 4, and the down is about 8.25 clo, closer to your calculation.

Lots of assumptions here, but it seems like we should be able to get within a factor of two.

I'd like to do this test myself — but I don't own any down garments!

Richard,

You never cease to amaze me! That "home test" is fantastic. As Harold has asked, if you could fill us in on the temperature of the hot plate and the ambient room temperature, that would further put this in perspective.

I also noted with interest the diffence between the well used vs. the new Primaloft garments.

Keep up the good work and thanks!

Harold – All measurements were in F and the ambient room temperature was 71.7F (thermastat was set to 70F). The test area was free from major radiant heat sources (windows & heating vents). The hot pad I used was the Sunbeam Sports Wrap with Intelligent Controller, Model 902. I used the high setting because it achieved equilibrium temp the fastest and regulated the best of the multiple settings on the heat pad. The following is a temperature time plot for the H setting.

The test geometry was the same for all tests. The garment was fluffed and the back was centered over the pad. The temperature sensor was also centered over the pad.

Ideally I would set the heat source for the average skin temperature of 93F. My heating pad provided 91.4F for the 3 setting and 101.3 for the 4 setting. The heat pad's temperature controller did not ramp up and stabilize quickly for either of those settings. The following is the 3 setting plot.

Daniel & Michael – Primaloft One is one of the worst synthetics for loosing loft / insulation value through stuffing and washing but, all synthetics exhibit that characteristic to various degrees. This is in stark contrast to down’s durability.

Polarguard Delta is one of the best synthetics for durability. Yet, I typically loose 30% loft from a new Polarguard Delta sleeping bag after one season of use. Granted… I use a compression sack to minimize packing space but, I only wash a synthetic sleeping bag once per season.

"The most striking observation to me is not so much the graph, but how much loss of insulative value the primaloft jacket suffered."

One thing to remeber when looking at this graph is the base isn't at zero. Looking quickly at the graph it seems the primaloft vest and jacket lost over half there insulating ability, but the difference is only from just over 81% of ideal to just under 79%. Since the temperature differential was 65 deg F, 2% is only about 1.3 degrees difference in thermometer readings.

Richard,

Thanks for the details of your home test. From your description ("Intelligent Control" and temperature graphs), it sounds like your heat pad may not output a constant power but instead regulate to a fixed temperature. If the power output varies by a lot, my estimates of relative thermal conductivity may not hold. Without knowing those properties of your heat pad, ultimately I don't know what to conclude, unfortunately.

The temperature differential between the garment surface and the ambient temperature may also give us some useful information about the heat flow, but I don't know enough about heat transfer in air to make any statements regarding that.

James,
Good point about the relative scale for insulation degradation. This seems to indicate that even if the garments lose loft (as Richard observed) the thermal properties are not greatly affected. Does this match with subjective experience?

My 1st down garment was The North Face's first heavy down parka in the mid '70s. I sold it right away after getting the back soaking wet while backpacking in -5 F temps. Went to an Alpine Designs (remember them?) synthetic parka & was very happy.

Down is Ok for camp use and ice fishing or promenading at Vail Village or your local haute coture store. REALLY good down (i.e. 800 fill plumules from mature birds) is necessary for extreme mountaineering due to weight savings.

Eric

That is a very cool test and a great starting point, but I think the differences in temperature from what we experience in the woods vs. ambient temp of a house and differences in hotplate temp and heat emitted by a sleeper may skew the results a bit. Here is my reasoning:

From what i gather, by comparing the fugu and the jacket/vest with 1.2" primaloft, the primaloft insulates about 90% as well as the 1.5" of down with 80% of the loft. Thats an interesting result which promotes synthetics being better per inch at warmth, but it seems a bit off kilter from 1.2" of primaloft making a 20* rated bag (15* depending on the companies ethics) as reported by several sources, and that also according to several highly esteemed companies, 2.5" of single side loft w/ high quality down makes for the same 20* bag. The difference there is over 2 times the warmth for the same loft as down. It is also similar for other modern synthetic insulations like climashield XP and combat. I would highly recommend people still relying on loft to measure synthetic bag warmth to read up on clo measurements to find the warmth of what they want to make or buy because not all inches are the same.

so if I've understood right, Richard, what you're saying is a given amount (weight) of down per volume unit is more thermally efficient when slightly compressed (by baffle height limit) than if left to loft fully. Is that correct?

And if we compare the best quality down with the best synthetics we should then be using this slightly compressed down for the comparison to be fair and down would still be warmer by around 30%, right?

Let me recall it's the warmth/weight ratio that we want to maximize, nevermind the loft (in case warmth and loft don't have a 1 to 1 relation, which is what's being discussed). I've read too many posts that discuss loft and warmth but don't mention weight. Maybe it's implicit data and I don't know but I guess I don't care if a less lofty synthetic insulates as much as a loftier down item if it still weights more.

Inaki,

I'm not Richard, but I'd like to chime in. I agree that the more important question is warmth/weight. However, the question of warmth/loft is important if we want to compare the warmth of a down garment and a synthetic garment, with the ultimate goal of then comparing the weights of equivalent warmth down and synthetic garments so we can evaluate the trade-offs.

From this thread, I feel there is no settled consensus on how to compare the clo values of a down garment with a synthetic garment of known fill material. A rough ballpark estimate could be made if we knew the relative warmth/loft of down, but there seem to be contradicting experiences and facts regarding this (Richard's quoted thermal conductivity of down vs Ayce's experience and David's observations on sleeping bag ratings).

Finally, in regard to how densely down should be packed, I tried to make the point above (not very well, I guess) that it still makes sense to pack down for optimum loft, the reason being that even though the "k" value Richard listed is higher (more heat conduction, worse), the overall insulating power (which goes as thickness/k) is better for the same amount of down, because the less densely packed down is twice as thick.

Yikes! My head is starting to hurt!

Are we saying that a 1" thick down garment, that is stuffed to allow the down to reach its natural loft, when compared to a 0.8" thick garment using the same weight of down, therefore compressed by 20%, that these two pieces would be approximately, equal in warmth and would weigh about the same, minus a little due to shorter baffle height?

I can't find it now, but there was an article by BPL about 3 years ago, comparing high loft down jackets. As I remember, the reviewer remarked that the Rab Neutrino (which is sewn thru) felt warmer than the jackets that had baffle construction. The reviewer attributed this to the Rab being over-stuffed. That article would lend credence to this idea.

Harold, your previous explanation was fine and your conclusion seemed logical to me. I know I may be missing something as I don't fully know the meaning of these K value or the units involved (ok, it's heat conduction, less is better… but I don't know if the variation is linear with the weight/thickness/density). Then I was expecting Richard to clarify the discrepancy and he answered your message but not this question so I was asking again.

Richard, you provided very interesting info in that answer to Harold but unless I'm missing something this question is yet open. And since it seems clear that higher density down has a better warmth/loft ratio, it seems pretty useless to even try to compare warmth/loft between down and synthetics if we don't settle on which down setting first.

I really hope what I'm saying makes sense and I'm not complicating things further. If I am, please tell me to shut up :)

Here's an experiment I did last weekend tetsing down vs synthetic bags.

Conditions: It snowed about 3 inches, of which 1-2 melted. We were by a steaming lake (probably 50F compared with the ambient temperatures of 35F in the day and 26F at night). Thus very humid and fairly cold but not deep cold that sucks the moisture out of the air.

Gear: For the weekend I used a dual quilt system: a custom 13oz Nunatak bag with 2-2.5" of loft under a Cocoon 180 with 1.5" loft. I used a bivy sack and camped in a Black Diamond Mega-light.

Procedure:
7:30pm–go to lake, remove pants, pile hat, pile sweater, pile gloves and submerge till soaked in lake. Wring out until no more water dripped from clothing but not as dry as it could have been with further wringing. Put clothing on over wool base layer and medium weight capilene. Remain at fire an hour while "dripping" a bit. Still quite wet at the end.

8:30pm–Removing clothing except light base layer, put vapor barried clothing on over the base layer, put all wet clothing back on. Socks now soaked from water dripping from pants. Go to bed. [Vapor barrier clothing used to help separate clothing from body moisture issues and to further test VB clothing in warmer conditions.]

In the morning clothing had mostly dried. They were still moist but comfortable–I changed out of the vapor barrier and back to the (previously) wet clothing. The night was comfortable but the bag was very moist.

At home I weighed the bags and measured their loft before and after drying.

RESULTS:
Nunatak down Cocoon 180
Pre-Drying
Weight 14.4oz 19.7 oz
Loft * 1.25" 1.125"

Post-Drying
Weight 13oz 17oz
Loft * 2.25" 1.5"

Loft Lost% 45% 25%
Weight Gained%10% 16%

*It is very difficult to accurately measure loft–I fluffed both bags, placed them on a wood floor, held a yardstick on the surface so most of it touched the bag (two measurements side to side, one top to bottom with the three results averaged) From here I pushed a ruler down vertically until it hit the floor (compressing the material) and read the measurement against bottom of the yardstick.

I wasn't surprised that the down lost so much loft. I was surprised that it lost so much loft despite having gained less water weight. The synthetic bag gained 50% more water weight, but lost only half loft of the down bag. Interestingly, even after a very wet night the down bag still weighed less and provided more loft than the synthetic. However with another night out in similar conditions the trend wasn't promising.

Based on this test as well as the difficulties keeping two quilts aligned and closed as I turned from side to side as I slept, I would bring a synthetic bag in conditions such as we faced on this trip. (moist/cold when clothing is likely to get wet).

I've had good results using vapor barrier clothing and down in the winter (below 30F) when clothing doesn't generally get wet. I'd like to experiment with using VB clothing in moist above freezing conditions when I'm sure I can keep my clothing dry.

Harold & Inaki – If you opt for 800 fill down baffles filled for optimal thermal density, they will out-perform (>oz/inch/yd2-green, >clo/in/yd2-red, >clo/oz/yd2-blue) the best synthetics when dry. If you opt for max loft only, you will have better oz/inch/yd2-green (eye candy) but you will not have a higher clo/oz which is the most critical measure for backpackers. I cannot quote you the clo/oz for the lowest density because I am not aware of any labs that have published W/m K values for sub optimal thermal down densities.

My simple relative performance heating pad test demonstrated the same relative results as this spreadsheet. The down garments are warmer than the new Primaloft One garments of comparable thickness (average down baffle height rather than max down baffle height).

I created the above spreadsheet for use in this post. Down is very complex to compare with synthetics because of the variables: down loft and down bulk density. I attempted to focus on just one representative down combination to simplify its comparison with synthetics.

Michael – The article you referred to was entitled, "Winter Backpacking Comfort: Lightweight Gear and Techniques for Shelter, Clothing, and Sleep Systems"

It said in part, …"The Rab Neutrino has overstuffed sewn-through baffles, a hood, and a full zip. Despite its sewn-through baffling, we found the Neutrino to be warmer than many other down jackets using baffled construction – a testament to the attention to detail and fit that Rab paid to the design of this jacket. For the winter alpinist, this jacket may offer the highest warmth:weight ratio of any insulating jacket available…"

High quality down compressed to the optimal bulk density, is unsurpassed in clo/oz when kept dry. The over simplified phrase "thickness equals warmth" is used because attempting to explain what really equals warmth almost always results in, "Yikes! My head is starting to hurt!”

My prior post with the spread sheet should answer your first question.

Inaki – Yes the variation is linear with thickness. If you double the thickness, the resistance is doubled and conductivity (K) is reduced to 1/2.

.

Harold-I multiplied by 2.5 because the product is the weight in oz. Look up the weight of a 1" thick yd2 piece of Primaloft to see the consistency in this green column. You will find that it weighs 6 oz and 1" thick yd2 Climashield XP weighs 4 oz. Everything in the green column is the weight of a 1" thick sample yd2. Please consider this information in place of your initial assumption and give it another try.

Maybe this graph will make the density factor easier to understand. For construction purposes a material is defined as insulating only if its thermal conductivity is significantly less than 0.06 W/m K. Uncompressed down would not be considered an insulator by most building code standards.

Another way of looking at max loft down with a thermal conductivity of .06 is that each inch of the eye candy uncompressed down equals the thermal insulation provided by the same thickness of dry snow or saw dust. Granted that down is LIGHTER than snow or saw dust but the thermal insulation value is abysmal.

Also note that Primaloft One, also known as Bimodal Polyester Artificial Down, is only offered in battings with a thermal conductivity value of ~.033. All of the less dense green dots above this point could be offered to the market but are not. Is it now obvious why not?

"Conditions: It snowed about 3 inches, of which 1-2 melted. We were by a steaming lake (probably 50F compared with the ambient temperatures of 35F in the day and 26F at night). Thus very humid and fairly cold but not deep cold that sucks the moisture out of the air.

Gear: For the weekend I used a dual quilt system: a custom 13oz Nunatak bag with 2-2.5" of loft under a Cocoon 180 with 1.5" loft. I used a bivy sack and camped in a Black Diamond Mega-light."

Interesting experiment. It would also be interesting to see what happened if each bag were used separately or if the Cocoon were used under the Nunatak, since whatever passed into the Cocoon in your experiment had to pass totally through the Nunatak.

Hi Richard,

Thanks. You are right; I was being an idiot (although you were nice enough not to say so). You have the density in the 4th and 5th columns. I'm going to edit my previous post so I don't drag others into my confusion.

I'd like to return to the question of clo/oz for the uncompressed down, though. The following from what I said in the post I edited still holds, I think:

When you say you do not know any values for W/mK for
suboptimal density down, you seem to ignore the first
graph you posted in this discussion. There you have .32
btu in/(hr ft^2 deg F). The conversion to W/(m K) is
multiplying by .144, so the max loft density down has
(granted, for "average down", and not 800 fill) a .046
W/mK and the "max efficiency" density down has
.036 W/m K. I thought it was reasonable to infer from
this that the optimal loft 800 fill down should have a
clo/in of 6.562*.036/.046 = 5.13 clo/in. Or do you
believe that 800 fill down has a markedly different
compression-dependence of thermal conductivity than the
"average" down so we cannot infer this?

If we use the value 5.13 clo/in for uncompressed down, we find that the clo/oz is 5.13 / 1.62 = 3.2 clo/oz, still higher than the compressed down.

In the last graph you posted you also have data on thermal conductivity vs down density. If you present that data on a scale where we can see the bulk density for each data point a little more clearly, we can find the optimum density by maximizing the quantity
1/( [Thermal conductivity (W/mK)]*[density kg/m^3] ), which is proportional to clo/oz. Or you can just plot that quantity vs density. These data may show that there should be some compression to optimize clo/oz, but the data in your original chart (with the points at .32 btu in/hr yd^2 deg F for .25 lb/ft^3 down and .25 btu in/hr yd^2 deg F for .5 lb/ft^3 down) clearly shows the least compressed down has the best clo/oz.

We can do the same thing for the bimodal polyester (PL one) data. Then we can find out if the .033 w/mK batting has optimized for clo/oz.

Incidentally those data also tell us the quality of down measured — for example 800 fill down at max loft has a density of 2.16 kg/m^3. The first graph you posted had a minimum density of .25 lb/ft^3 which corresponds to 430 fill down.

I don't think that what building code standards require are necessarily applicable, because they care about how thick the insulation is (presumably because it takes extra material to build thick walls to hold that insulation). Hence thermal conductivity in W/mK, or the resistance clo/in (both are normalized to thickness) is important. In garments and sleeping bags the thickness constraint is not as important (because extra thickness is just making wider baffles), and we care about clo/oz (normalized to weight).

I had been meaning to do this after my last post, but was traveling yesterday. Using the data from the graph in Richard's post on 09/24/2007 23:08:15 MDT, I have plotted the clo/oz values for down at various compression. From the minimum density, I estimate this down to be 430 fill.

The least compressed down has the highest clo/oz.

The numbers used are in the following table:

In this table, columns 1 and 3 are read off of Richard's graph. Column 2 is derived from column 1, while columns 4 and 5 are derived from column 3. Column 6 is column 5 divided by column 2.