Quantifying the risk for thermal excursions on transportation lanes in temperature-controlled logistics
A key challenge in temperature-controlled logistics is finding a container to keep the payload at its allowed temperature along specific transportation lanes. We use thermal simulation to predict the performance of any container along any lane, based on accurate container models, and temperature data from about 20’000 weather stations worldwide.
Classic approach: physical testing
Classically, attempts have been made to quantify lane risk by making test shipments, or running tests in a climate chamber, both very costly and time-intensive processes. If the test shipment fails, you have to change the container and run another series of tests. If the test succeeds, the implications are of limited use: the result gives you a performance snapshot of only a single day out of many different possible transportation scenarios. You would have to run many tests on different days and seasons to get a viable lane-specific characterization of the container’s thermal performance. For another lane you would have to do similar tests all over again, or resort to reference profiles. This is a very time and money consuming process, which still leaves you with great uncertainties.
the result gives you a performance snapshot of only a single day out of many different possible transportation scenarios
Our approach: thermal simulation
Our virtual cold chain approach is an effective tool to find the best solution for a transportation lane through robust quantification of the lane risk. We prepare virtual models of containers and expose them in thermal simulation to air temperature data characteristic of the assessed transportation lanes. Since thermal simulation is very fast, it is easily possible to assess thousands of lane temperature scenarios, corresponding to years of test data, within less than a single day. The results can be used to compare the thermal performance of different containers on a lane, or to test the suitability of a particular container on different lanes. Being cheap and fast, this approach invites you to explore all parameters to minimize the lane risk: vary the container and its materials, change lane parameters like locations and starting times, and so on.
Let’s find the best solution for shipping temperature-sensitive pharmaceutical products at 2-8°C along two transportation lanes, (i) Dusseldorf via Atlanta to Chicago, and (ii) Chicago via Miami to Sao Paulo. We need to decide between three available container packouts: summer, winter and all season. The question is: do we use a summer and a winter seasonal packout, or do we use a single all season packout? If we use the seasonal packouts, at what time of the year do we need to switch between the two? The virtual models of the three packouts are shown in the following figure. The payload is the same for all.
Figure 1: Three container packouts winter, summer, and all season. Winter and summer packouts use a hull from PU foam, and water bricks preconditioned at +5 and -5°C. The all season packout has a hull of EPS, and PCM bricks with an initial temperature of 5°C, and a phase change range of 4-7°C.
Temperature profile of the lane
The second piece of information we need for quantifying lane risk is the ambient thermal condition on the transportation lane. We get this information from historical temperature records of weather stations along the lane. In the figure below we show the ambient air temperature profiles along the two example lanes. Each gray line represents a single day during the time between 2011 and 2016, a total of about 2000 individual temperature profiles per lane. During flight intervals, the ambient temperature is assumed to be controlled and set to 15°C.
Figure 2: Temperature profile of the lanes Dusseldorf – Atlanta – Chicago (top) and Chicago – Miami – Sao Paulo (bottom). Each gray line represent a single day within 2011-2016. Data from historical records. Flight intervals are assumed controlled temperature at 15°C or 20°C. Blue and red envelope indicate minimum and maximum profile, respectively.
Determine the risk for thermal excursions on a lane
With these two pieces of information, container and lane, at your fingertips, you are only a few clicks away from results of the lane risk analysis. Because simulation is fast, it is no problem to run each and every temperature scenario of the selected time interval. The result is shown in the next figure, compiling the overall thermal performance of each container packout for each lane on every day during the years 2011-2016.
Figure 3: Lane risk quantification for three container packouts on two lanes. On lane ORD-MIA-GRU, the all-season packout performs best. On lane DUS-ATL-ORD, a combination of all-season packout and summer packout performs best
For the lane ORD-MIA-GRU, best performance is obtained using only the single all-season packout for the entire year. For the lane DUS-ATL-ORD we can see that a single packout is not sufficient. Deeper analytics show that we get best performance by switching between summer packout during the months May-Sep, and all-season packout during the months Oct-Apr.
Lane characteristic thermal performance indicator
The kind of analysis shown here has not been possible in the past using classical approaches. To collect the data required for this analysis, it would be necessary to combine data from about 12’000 individual climate chamber tests: (#boxes) x (#lanes) x (#profiles), thus 3 x 2 x 2000 ~ 12’000. Since each test takes about 2 days, this would lead to a total required testing time of about 65 years – an unfeasible task, even if multiple parallel chamber tests were carried out. With our simulation approach, assessing all 12’000 scenarios takes only a single day, providing a statistically solid, lane-characteristic, packout performance indicator.
Of course, simulation is not restricted to the examples shown here. You can run analogous simulations for any packaging and any lane. Not only will this approach save money and time, it will also make your prequalification process a lot more transparent and add a sound scientific basis to your GDP compliance requirements.
Share this article