Yes, I agree that it doesn’t seem to make sense. Also, look at how the bottle temp cools slightly before the ambient temp (on its way down) even reaches the bottle temp. Doesn’t seem to make sense.
I can tell you that I tracked both sensors in air for about a week prior to this experiment and they were dead-on with each other (and with a mercury thermometer that I used to verify that the sensors were accurate).
Yes, I’ll try to post a zoomed-in view of the data shortly.
Paul, great stuff, thank you!
My experience living 75 miles south of satan’s blast furnace makes shipping a no-go nine months of the year. I’ve found that using the styro shippers with the cold pack works fine for overnight from LA wine shops. I do not hesitate to order a case of rose for immediate consumption during the summer here. I don’t ship anything else though. I’ll wait for cooler days in Nov, Dec and Jan for anything valuable or destined for the cellar. ![[cheers.gif]](/uploads/db3686/original/2X/0/0ff9bfcdb0964982cd3240b6159868fbdf215b1a.gif "cheers"){kind=small}
Very cool and thanks for sharing the pictures. Post of the month!
Here’s a 2 day zoom-in of the chart after it had stabilized a bit:
My earlier comment about the liquid cooling before the ambient temp got below its temp is false - it just appeared that way on the wider chart. It’s now clear that this is not happening.
You know if you keep posting useful stuff it is going to make the rest of us look bad. Or worse.
Some empiric data; In 2012 I drove my wine up from Texas to Virginia. It was all packed in various boxes–moslyt styro but some pulp and started at cellar temp. It was a 3 day drive and temps were pretty consistent. I had an ambient temp monitor in the U-Haul truck and another wireless monitor in a styro box. Generally, lows at night were in the 50s, highs in the upper 70’s. In the styro box temps were more less extreme, more like 60-70. I did all kinds of crazy things when it got warm/sunny like pack the truck with dry ice and park in the shade and wait till dark to drive, etc. When I got there ambient temps were 78! Styro box ~74, but wine bottle clearly cooler, probably high 60’s. Wine in pulp boxes felt room temp(but none felt warm). I took this to mean after ~72hrs the pulp boxes had nearly equilibrated with the ambient temp while the styro hadn’t yet. I think the op’s post would support this conclusion with the styro shippers.
Thanks for real data!
-jd
Thanks for the data keep it coming. Specifically I wonder how much better the wine would fair if it was a full case with 12x the thermal mass of liquid and glass.
Great work, Paul! This somewhat confirms what I’ve always thought - I prefer the styro shippers. I know they’re the least environmentally friendly, so it would be great if there was a way to re-use them.
Counter intuitive, I know, but most evidence (more anecdotal but there are some tests on wine biz monthly) says this is a wrong assumption…styrofoam is not the insulator people think it is when its comes to wine shipping. Takes a bit longer to heat up, yes, but also longer to cool down. I’m no expert, but I think that over the course of 24 hours the faster cooling is more of an advantage over slower heating. I’ve had this conversation many times over the years with FedEx and with big DTC fulfillment providers and everyone says same thing…they’ve done the tests and pulp performs better.
Can we run this experiment with styro and pulp shippers with same thickness of outer shipper?
Cameron,
Interesting point. But it would seem that for a given ambient outside temperature (heat source) the maximum temperature reached by the bottle (heat sink) would be higher in pulp, thus more likely to reach a critical affecting temperature. That would seem to be the most important attribute of a shipper, keeping the wine from reaching a damaging temperature. It would be interesting to see this test side by side for sure.
Hmm. It is a mistake to believe that cooling is faster than heating, that’s not true. All that matters is the temperature difference between outside and inside - the greater that difference, the faster the temperature change of the contents (you can see that by observing that the direction of temperature change inside the bottle turns over when the outside and inside temps are the same). You have more or less the extremes of this experiment right here: no insulation vs. the best insulation (in terms of what’s available for wine, that is). Anything else (like pulp) is going to be somewhere between those two. That means that in this case, pulp would give higher “highs”, which is the thing you really want to avoid.
I’ve wondered why the inside temperature appears to “average” lower than the outside temp (which is what I think people may be interpreting to mean that “cooling is faster than heating”). It may be that the two temperature monitors aren’t calibrated correctly. I’d like to see the result with the two monitors swapped inside/outside.
To fully understand why we see what we see on this graph, do we need a sensor in the airspace inside the styro as well? If the air in there does not get very warm, then the wine will be responding to the high/low of that, not the outside air.
These are awesome data!
Another interesting variation on this experiment would to examine different shipper sizes/configurations. Since the surface area to volume ratio is rather different between 1x3, 2x2, 3x4 shippers (etc) the ratio of (potential) heat flux to thermal mass would also be different. Add in the various styles of pulp vs styro liners and we’ve got a whole lot of configurations here…
Paul, fabulous data! I suspect the bottle probe may be reading a couple of degrees lower than the air probe.
Also it’s possible there is a difference in the response time of each probe. I remember once purchasing a combined temp/humidity probe, and when I put it in my active cellar it took it more than an hour to cool down and read the cellar temp. By comparison a temp only probe from the same manufacturer cooled to cellar temp in less than a minute.
It might be instructive to repeat the experiment but with both probes side by side in air. The result might help explain what’s going on in your graphs.
Yep, this would give the calibration data needed to correct any relative error.
Just for grins, here’s a simulation of something like the second half of Paul’s data, with just a constant diurnal temperature change between 70 min and 87 max. As expected, the temperature eventually equilibrates around the average daily temperature, which is what I’d expect Paul’s measured data to look like if both temperature sensors were calibrated.
Alan, out of curiosity, did you perform the same test as the OP with different temperature ranges?
It’s early, and my brain hasn’t fired up yet. Forgive my ignorance…
Drew, note that my graphs is just a simulation, not an actual measurement. So this is the theoretical behavior with a thermal conductivity picked empirically to approximate Paul’s results. I could run this with any time varying temperature graph, and any conductivity value. The simple daily sinusoid is trivial to set up, so I did that for this example.
You would think, right? But for some reason that doesn’t seem to be happening. As I said, I’ve had plenty of discussions with the big DTC fulfillment companies and they believe at best its wash between the two materials based on their numerous informal tests. All have told me they actually think the pulp performed better.
So, again, I am no expert but there is something going on here we don’t understand. I would love to see this test performed with pulp and styrofoam in the same environment. Wine Business Monthly did this about 7 years ago with results similar to what the was reported above. They were surprised as well to find that pulp outperformed styrofoam in their admittedly informal test Informal Test of Wine Shippers