Applicability of Human Thermophysiological Model for Prediction of Thermal Strain in PPE

Round 1

Reviewer 1 Report

Comments on Applicability of Human Thermophysiological Model for Prediction of Thermal Strain in PPE

The study describes development and validation of a method for deciding over exposure limits. The method is of interest to the users as it utilizes advanced modelling as a basis to create more simple tool with reasonable goo prediction capacity that can be affordable and usable also for users that can’t afford or don’t have experience, capacity or time to user the original advanced termo-physiological model.

Most of the comments are of editorial nature. I would suggest that the text is read by authors‘ colleague with fluent English or an English proof-reader, this is more relevant to the introduction and methods sections. When the authors come into results and discussion, then it is more clear.

Authors’ list: I suggest not to half a name at the end of line. Preferably there could be added a line break (Shift+Enter) just before Vladimíra Fialová

In affiliations e-mail address of only one (first) person of the organisation is given. There are 4 persons from one and 3 from the other organization in total.

L34: replace “smoked” with “smoke filled” or “smoky” (“smoked” is used in smoked meat, smoked sausage etc.)

L34 and L36: You mention here psychological burden. Do you mean psychological or physiological? Your study focuses on physiological measurements, and there are quite some related references there in the text. However, if you still mean psychological burden here, then it would be great if you add here also some reference to relevant work on PPE and psychological effects.

L46, 72.3 % of respondents to perceived ‘hot’: I am not sure if “to” here is the correct word or makes this part of the sentence confusing. Do you mean “who” here instead?

L56: Is the use of word “attacking” in “attacking 37 °C” the best choice of words? Do you mean reaching 37 °C or exceeding 37 °C?

L63 & reference 6: Check the reference. It refers to the contributor list and not to a specific chapter. If you refer to the whole book, then refer to the book title and the editors. If you refer to different chapters, then give a separate reference for each.

L82-85: References 19 and 20 are missing here. They occur later (L96 and L98). The references may need to be renumbered from 19 and up to and including reference 24.

L150, L153, relations/relation: equations/equation

L156 & reference 28: Differently from other references to standards, reference 28 starts with “technical committee …”. Start with ISO 9920. Technical committee number is not needed.

L170: Ta has already been explained at Eq. 1.

L180: Eq. 3 does not include Hc. Also, it is already explained at Eq. 1.

L190: Here you describe human tests. Both in L193 and in results (L391) you refer to VO2max as a measure of fitness. I suggests to add a table with your test person’s parameters.

L192: Only at Table 5 it becomes clear what you mean by ”up to 12”. Make it clear already here, in methods that in most of the  exposures you used intentionally less than (but up to) 12 volunteers. This could also be a part of schematics as suggested instead of table 3, or a separate table or a table connected to test persons parameters.

L205, HRlim=220-age: I wonder why did you use this estimated, widely variating and individually less accurate limit, when you measured VO2max? I would expect that during VO2max measurements you also acquired real HRmax for each individual. The limit could be set just as a percentage of that value, e.g. 90 or 95 % of individual HRmax.

L209-L222: Some information is duplicated, while it is not clear what was measured with what and why some parameters were measured in multiple instruments. Are there some differences in instruments that could introduce systematic differences between the conditions or subjects?

L236; PHS The validity range of the model …:  The validity range of the PHS model …

L264 and L274: Stick to the same term for equations (relation, formula). I would suggest using “equation”.

L111 vs. L282: Use the same terminology for the tool throughout the text – either Predictor of the Thermal Stress or Predicted Thermal Stress.

L286 & L534: What is GUI?

L290-L292: Unclear sentence , especially the second half of it. Is “can” a right word there or is it in the right place on L292 (creates most confusion)?

L297: Table 2 and Table 3

L305 with Trad in Table 2 vs. L251 with Tr in Table 1: use the same abbreviation for the same parameter. If Tr and Trad differ, please give definition for both. Check if all other parameters use always the same abbreviation.

Table 3 and L318: Do you by “interrupted work” mean intermittent activity? Can you provide schematic of the work protocols, possibly instead of Table 3?

L324-L325, Check the sentence. I believe understanding the meaning, but the sentence composition, word order/use is somewhat weird. Could it be instead: “The defined garments, including their local characteristics, were entered into the FMTK model”?

L343-L344: This sentence repeats the statement of L341, doesn’t it? If there is difference of the meaning there then make it clear, if not, then eliminate one of the statements.

L351, L354: relations – equations?

Figure 2. What novel information is it bearing compared to Figure 1? Could you give a subtitle for each picture that describes what is there? If you want to show the manikin and the used underwear (not visible in Figure 1), then could the 1^st and the 4^th picture be enough (one without and other with skin)?

L349 & L351 vs. Figure 3: The text mentions total thermal insulation of clothing while the figure shows only the local values. Would it be proper to refer to Table 4 instead? When mentioning here the evaporative resistance ReT (L351), then the reference to Table 4 would also be proper. Possibly Table 4 should be available before Figure 3.

Figures3 and 4: did you take into account that during the subject tests with SCBA there will be available considerable face cooling and removal of humidity under the mask that was not present in manikin tests?

L374: I suggest using “section” instead of “chapter”

Figure 5 generally repeats information in Figure 1 (or vice versa – more information is available in Figure 5). I wonder if all these figures (also Figure 2) are needed? If to replace Table 3 with a schematic of experimental procedures, then possibly some pictures would fit to illustrate that.

L394, Tre,0: was it controlled what the subjects did before experimental and how they arrived to the laboratory, e.g. by car, bus, bicycle, being in hurry and running, being early and waiting in the lobby? Did you use pre-conditioning time, e.g. 30-60 minutes to relax, if they had been physically active just before?

L399: a fever reaching above 38.1-38.2 °C is commonly feeling the worst, while if it goes above 38.7-38.8 it does not have such adverse psychological impact. How well the subjects were trained or acclimated? Repeated training where core temperature reaches above 38.5 °C, helps to go over the thresholds available in body to protect from reaching dangerous temperatures. Well-trained marathon runners core temperature can reach above 40 °C (a classical sports study by Nielsen et al).

Section 3.3 in relation to PHS: This comment is not directly related to your results but to the use of PHS in general. PHS is not meant to be used with impermeable protective clothing, and to my knowledge, is not validated for this range. I wonder if PHS prediction would have been closer to subjects data if evaporative resistance of clothing had been kept at the top end of the allowed range and relative humidity of the air would have been increased to the level that would allow only minimal evaporation, that would correspond to the expected evaporative heat loss (generally something related to clothing ventilation through openings during walking and heat pump effect, e.g. see Havenith et al. 2008 & 2013). See also your similar suggestion on temperature adjustment in L726-729.

L509, 116644: Do you mean 11664 scenarios?

L528, Individualized: in Figures you use individualised with “s”. Use it the same way also here.

Section 3.5: Your simulation results in PTS are based on the conditions and clothing properties that you actually measured. Do you expect the outcome to be as good if completely different protective gear and conditions were used? I.e. is there any bias in the validation as you performed here? Partly you address the issue with the last paragraph of this section but it is still about the tested clothing …

L623, corresponding: corresponding

Discussion (and conclusions): The discussion is well focusing on your results. I wonder if any references to other publications may strengthen your statements? This comment does not insist adding them but I believe that adding some relevant ones can be useful.

L676-682 (see also comment on L399 above): I agree with the discussion. However, it should be remembered another risk if core temperature reaches 39 °C and above. When above 38-38.2 °C many people start feeling bad and want to quit or slow down work rate (natural body signal to protect the organism from adverse effects of overheating and an alarm to take the measures), then at core temperature of 39 °C and above the body alarm threshold has been breached and the feeling bad sensation disappears (compare with the sensations during high and very high fever). This creates additional risk for persons who have note been trained to notice other signs of body to counteract overheating. In these cases individual supervision by someone in charge is essential (another person who orders to stop the exposure). If the PTS tool may allow for individual increase of exposure according to subjective sensation, then considering the physiological responses described as above, a cut in exposure must be introduced irrespective what the individual feels. Only if a model has been adjusted and calibrated individually, and the person’s reactions/limits/physiological behaviour are known, then limits set for average or even specific population may be exceeded, e.g. setting individual criteria for top athletes.

 

Last 2 paragraphs of discussion and conclusions: planning can also involve being ready with sufficient water supply, expected number of emergency measures/equipment for body cooling, or equipment to create cooled shelter or shade from solar radiation for breaks and recovery.

See the comments in the previous section. 

Author Response

Dear Sir/Madam,

I would like to thank You for the thorough review of our article.

We incorporated Your suggestions as well as native English speaker modification into the new version of the text.

In addition to majority of the comments incorporated directly to the text I would like to answer some of Your questions:

L205, HRlim=220-age: I wonder why did you use this estimated, widely variating and individually less accurate limit, when you measured VO2max? I would expect that during VO2max measurements you also acquired real HRmax for each individual. The limit could be set just as a percentage of that value, e.g. 90 or 95 % of individual HRmax.

We followed the standard test performed in our laboratory for long time with the limit HRlim = 220 – age. But, obviously, we can see that the accurate measurement using VO2max would be better solution. However, we wanted to finish the whole set of the tests with same limits given at the beginning of the project.

Table 3 and L318: Do you by “interrupted work” mean intermittent activity? Can you provide schematic of the work protocols, possibly instead of Table 3?

We changed the term, but we would prefer to keep the table form of the general Work regimes, because the schema might become rather complicated as it is possible to adjust particular metabolic rate for each work regime.

Figures 3 and 4: Did you take into account that during the subject tests with SCBA there will be available considerable face cooling and removal of humidity under the mask that was not present in manikin tests?

Yes, we are aware of the fact and it was observed by the probands during the real tests, but it was not considered as a further variable/boundary condition for the verification of the thermophysiological models.  

L399: A fever reaching above 38.1-38.2 °C is commonly feeling the worst, while if it goes above 38.7-38.8 it does not have such adverse psychological impact. How well the subjects were trained or acclimated? Repeated training where core temperature reaches above 38.5 °C, helps to go over the thresholds available in body to protect from reaching dangerous temperatures. Well-trained marathon runners core temperature can reach above 40 °C (a classical sports study by Nielsen et al).

We followed on the laboratory standard protocol with max RT limit 38.5 °C. The true is, that RT temperatures above 38.3 °C are for most probands very unpleasant. However, some of them are able/responsive to stand up to 38.5 °C, one proband (previous marathon runner) feels fine even at 38.5 °C, but we always quit the tests when getting up to 38.5 °C threshold. Our probands are not active young sportsmen so we have not tried to follow with the tests over 38.5-38.6 °C. Very often, the RT rises up to 38.6 °C or even 38.7 °C after termination the test. We keep the protocol – maximum 1-2 tests per week with one proband. We assume, that, in reality, using PPE against CBRN agents supposed to be very occasional and, moreover, the user has to be expert in CBRN agents handling, so we prefer to test with probands of „ordinary“ fitness, not the active sportsmen performing regular strenuous training. Our probands group consist of researchers, voluntary firemen, nurse, policeman… mostly with „normal“ fitness level. 

The true is, that in current research, we have cooperated with young active athletes as probands and 1-2 of them (from the group of 6) were fine at the temperature above 38.5 °C.  But 2-3 of them were exhaused around 38.3 °C also despite their young age and excellent fitness.

Section 3.3 in relation to PHS: This comment is not directly related to your results but to the use of PHS in general. PHS is not meant to be used with impermeable protective clothing, and to my knowledge, is not validated for this range. I wonder if PHS prediction would have been closer to subjects data if evaporative resistance of clothing had been kept at the top end of the allowed range and relative humidity of the air would have been increased to the level that would allow only minimal evaporation, that would correspond to the expected evaporative heat loss (generally something related to clothing ventilation through openings during walking and heat pump effect, e.g. see Havenith et al. 2008 & 2013). See also your similar suggestion on temperature adjustment in L726-729.

We added a sentence highlighting that the tests were outside the validity of the PHS and FMTK model aiming to prove if the model are still usable or not. We follow up with some other tests with other PPE ensembles and we would like to perform similar validations of PHS and FMTK models for the different PPEs in near future. Conducting the tests under different climatic conditions (higher humidity) would be also very helpful, but, it depends on other circumstances if it would be possible to follow with this direction of the research.

Section 3.5: Your simulation results in PTS are based on the conditions and clothing properties that you actually measured. Do you expect the outcome to be as good if completely different protective gear and conditions were used? I.e. is there any bias in the validation as you performed here? Partly you address the issue with the last paragraph of this section but it is still about the tested clothing …

As I mentioned in the previous answer, we have conducted set of tests with different gear and we want to prove both models in similar manner as described in this paper. However, we would like to publish the corrent set of the results.

L676-682 (see also comment on L399 above): I agree with the discussion. However, it should be remembered another risk if core temperature reaches 39 °C and above….

As I mentioned in the answer for L399. Moreover, our aim with the PTS tool is to include it in the firemen education, training and exercises to learn (by means of real on-site monitoring) the trainees how to subjectively realize the signs of threating overheating. And compare the PTS estimation with their subjective feeling. As always, this follow-up research depends on the concern of the stakeholders and related resources. Up to our knowledge, there is high concern among ordinary first responders regarding risk of heat stress in PPE and related possibilities for reducing the heat risk.

Author Response File: Author Response.pdf

Reviewer 2 Report

In thisr study, authors have documented investigation necessary for development of new user-friendly computational tool Predictor of the Thermal Stress (PTS) for the prediction  of thermal stress and permissible duration of activity in PPE. The work is original. The scientific quality of article is quite high. In my opinion, this review can be published in this form.

Author Response

Dear Sir/Madam,

I would like to thank You for the thorough review of our article.

Reviewer 3 Report

The manuscripts aims to study the applicability of the thermophysiological models for prediction of thermal strain and permissible working time in contaminated environment. The relationship between the thermal-insulation characteristics of four types of PPE and the induced thermal strain in set of real physiologic strain tests with human wearing PPE in a climatic chamber. Two thermophysiological models PHS and FMTK were analyzed with the experimental data and PTS was used to provide user-friendly platform for estimation of thermal stress in PPE.

The mythology is good to have a testing to measure thermal resistance and it is innovative. It can be accepted after following modifications:

1) The introduction is too redundant and should be abbreviated, such as Section in Page 3, "An example...."?

2)In Table 5, it has good repeatability for temperature? Actually, it is greatly important and key technology in developing manikin. So, the description of the tester should be added?

3) It can be seen from Figure 5 that the enviromental condition is variable, how do you guarantee the stable condition. It is very important in developing the tester?

It is redundant and can be improved.

Author Response

Dear Sir/Madam,

I would like to thank You for the thorough review of our article.

We incorporated Your suggestions as well as native English speaker modification into the new version of the text.

We incorporated Your comments straight to the text:

1) The introduction is too redundant and should be abbreviated, such as Section in Page 3, "An example...."?

We shortened the text of this article.

2)In Table 5, it has good repeatability for temperature? Actually, it is greatly important and key technology in developing manikin. So, the description of the tester should be added?

3) It can be seen from Figure 5 that the enviromental condition is variable, how do you guarantee the stable condition. It is very important in developing the tester?

We added more detailed description of the monitoring devices into the section 2.2 and additional reference regarding the uncertainty of the measurement (Fojtlín et al.)

The tests of  physiological strain are performed in the climatic chamber of National Institute for NBC Protection with regulated conditions (temperature, rel. humidity, air velocity), the actual values are monitored using multifunctional thermometer Testo 435-4 as added to Section 2.2.

The tests with thermal manikin Newton are performed in climatic chamber of Brno University of Technology with regulated conditions (temperature, rel. humidity, air velocity), see more information on the web:

https://eu.fme.vutbr.cz/about-us-dept-of-taee-laboratories-laboratory-thermal-comfort-2j

https://eu.fme.vutbr.cz/about-us-dept-of-taee-laboratories-climate-chamber-2j

More information regarding the used measurement method can be find in articles:

FOJTLÍN, M., J. FIŠER a M. JÍCHA. Determination of convective and radiative heat transfer coefficients using 34-zones thermal manikin: Uncertainty and reproducibility evaluation. Experimental Thermal and Fluid Science. 2016, roč. 77. DOI: 10.1016/j.expthermflusci.2016.04.015

TOMA, R. et al. Using a thermal manikin to determine evaporative resistance and thermal insulation – A comparison of methods. Journal of Industrial Textiles. 2020. DOI: 10.1177/1528083719900672

Round 2

Reviewer 1 Report

Thank you for improving the paper. It is now much easier to follow. I have still some comments, but considerably less than before. Two major ones on methodology at the start (take most space) are in large about the same thing and could possibly be clarified together or in an analogous way.

L47: … respondents perceiving ‘hot’, …

L79: e.g.

L237-245: It is still unclear why different measuring systems were used and for what, and how this might affect the results. The statement “The comparison of the systems is out of scope of this paper, see [30] for more details” is irrelevant from the viewpoint of your study. It would be understandable if one system was used, for example, skin and other for clothing temperatures; or if you state instead that [30] did demonstrate negligible differences between the results from these instruments, and thus, they were used as they were in order to reach the needed number of measuring points. Was the data from the backup instruments used? If not, then you don’t need to mention them. If yes, then it should be clear when rectal and intestinal temperatures may differ considerably. In the case of latter ones, it is also important to describe the procedures and timing of telemetric bill ingestion, and procedures before exposure, e.g. drinking of hot or cold beverages or nothing before the exposure, as there can be difference in measurements if the bill is swallowed 1 to 6 hours before. NB! In this respect regarding L258: myTemp does not measure rectal temperature and CORE does only the estimation of core temperature. None of them, in my opinion, is good enough for validation of rectal temperature-based model. I understand that you have your reasons for doing it as you did, but please motivate why it is not expected to affect your results. In L437-438 you mention that “considerable variability of results is observed in various probands even under the same test conditions”. Make me understand that it was not caused by use of different equipment or large variation in set conditions (see the next comment).

L250-251: ±1 °C and ±5 % from the set values is quite large. Is it reflecting standard deviation (SD) or amplitude? Amplitude of 2 °C is already high, but SD would be even worse. The stability of the test environment does not meet the standards. Were the maximum differences within individual exposures or for all experiments within one set temperature? What was the frequency of the environmental parameters’ change in the test chamber? I understand that you have the chamber that you have – I have for some time worked with a similar one and tried all means to make it more stable or adjust my tests so that the variation did not affect the outcome. Can you clarify why such substantial changes are irrelevant to your purposes? I don’t want a reader to assume that such variation is just always OK.

L273-275: May be repeatability of the subject tests is beyond the scope of this paper, but it may reflect the stability of the environment in you chamber, and the reliability of the results for modelling, and this is already affecting the scope of this paper.

L281: … both result in …

L282: … properties outside validity range …

L290: … is based …

L299-300: Both model … were …, however, only publications referring to Fiala model are referred at the end of the sentence. Should another reference be added?

Table 1, Range of validity of PHS and FMTK models [20, 30] …: Here are the references to FMTK, but not for PHS. If the text mentions PHS (and FMTK) then shouldn’t there be also given references to PHS either straight after it or in the same brackets? Is 30 meant to be 31 (30 seems not to be about the models)?

L348-349, … metabolic heat production (5 variants) …: It should be 6 variants according to text in Table 2.

Table 4: Sorry that I missed it before, but if ReT of the 2 first clothing sets (TF & OPCH) with merino sweatshirt (b) was not measured, then where you got im values?

L435, Figure 5: Do you mean Figure 1 and/or 2?

Figure 6 legend is on the next page. Move a paragraph before Figure 6 (L499-503) to after it.

Figure 10b shows an ambient temperature of 24,0, while legend (L699) mentions 25 °C. Real input is 25 °C. Probably in L699 Ta should be 24 °C as input value (L705 explains that it was corrected to 25).

L613, … on average at 38.2 °C in average …: delete “in average” at the end of the line. (By the way, Thermal Work Limit (TWL) suggests using limiting core temperature of 38.2 °C, e.g. see Miller and Bates 2007).

L722: … would have to quit …

L723: See the comment on L613 – your work seems also supporting TWL of 38.2 °C, while your warning in L764-766 may support 38.0 °C to be on safe side (as is ILO/WHO recommendation).

 

L747, … under the same other conditions …: Do you mean … under the same conditions …  or … under the same or other conditions …

English is now fine and there were just a few comments on some tabs.

Author Response

Answer to reviewer 1 for review 2

Dear Sir/Madam

Thank you very much for thorough comments, we hope it will help to better comprehensibility of the text as we, as authors, cannot detect.  

Moreover, we deleted the figures from the climatic chamber, as suggested in previous review as we deleted the description of backup monitoring systems. Then I had to move 2-3 paragraphs in few cases to avoid blank spaces on the page.

The minor comments are corrected directly in the text.

L237-245: It is still unclear why different measuring systems were used and for what, and how this might affect the results. The statement “The comparison of the systems is out of scope of this paper, see [30] for more details” is irrelevant from the viewpoint of your study. It would be understandable if one system was used, for example, skin and other for clothing temperatures; or if you state instead that [30] did demonstrate negligible differences between the results from these instruments, and thus, they were used as they were in order to reach the needed number of measuring points. Was the data from the backup instruments used? If not, then you don’t need to mention them. If yes, then it should be clear when rectal and intestinal temperatures may differ considerably. In the case of latter ones, it is also important to describe the procedures and timing of telemetric bill ingestion, and procedures before exposure, e.g. drinking of hot or cold beverages or nothing before the exposure, as there can be difference in measurements if the bill is swallowed 1 to 6 hours before. NB! In this respect regarding L258: myTemp does not measure rectal temperature and CORE does only the estimation of core temperature. None of them, in my opinion, is good enough for validation of rectal temperature-based model. I understand that you have your reasons for doing it as you did, but please motivate why it is not expected to affect your results. In L437-438 you mention that “considerable variability of results is observed in various probands even under the same test conditions”. Make me understand that it was not caused by use of different equipment or large variation in set conditions (see the next comment).

We left just the principal systems Viridia (HR and Tre) and Almemo (Tsk). The other systems have been deleted in order not to confuse the reader. Those reliable wired systems Viridia/Almemo were used as basic measurement for all tests, moreover we compared it to newer biotelemetric systems aimed for more comfortable usage in some of the tests. It is true, that the values measured by the biotelemetric systems were not used in the context of this article so deleting is reasonable option and it would be worthy to use the results for another article focused on comparison of various systems.

L250-251: ±1 °C and ±5 % from the set values is quite large. Is it reflecting standard deviation (SD) or amplitude? Amplitude of 2 °C is already high, but SD would be even worse. The stability of the test environment does not meet the standards. Were the maximum differences within individual exposures or for all experiments within one set temperature? What was the frequency of the environmental parameters’ change in the test chamber? I understand that you have the chamber that you have – I have for some time worked with a similar one and tried all means to make it more stable or adjust my tests so that the variation did not affect the outcome. Can you clarify why such substantial changes are irrelevant to your purposes? I don’t want a reader to assume that such variation is just always OK.

We put more explanations regarding the controlling and monitoring systems of the climatic chamber to the text. The stability of the environment in the climatic chamber is influenced by the accuracy of the controlling system (±0.4 °C an ±2.5%) and some inertia and time delay in dozens of seconds or a minute until the warm water gets to the exchanger and influence the temperature inside the chamber. The conditions were always set on required conditions of temperature (-10, 5, 25, 35 °C and 25 % rh) for all the tests, so there is, by my opinion, no reason to expect that the differences in results within  different tests we caused by variation of the climatic chamber. Up to our experience, such minor fluctuating changes in the temperature/relative humidity are not as significant regarding the results of physiologic strain tests comparing to other parameters – namely PPE characteristics and fitness level and acclimatization of the proband, but also actual condition of the proband including current fitness condition, menstrual period (for women) relating to fatigue/mood/enthusiasm, year season relating to acclimatization to the heat etc.

L273-275: May be repeatability of the subject tests is beyond the scope of this paper, but it may reflect the stability of the environment in you chamber, and the reliability of the results for modelling, and this is already affecting the scope of this paper.

As mentioned in the previous answer, the termination time of the test and physiologic strain level measured for particular proband depended on many various standpoints. For this reason we worked with a group of probands. In this study we calculated each person in the modelling just once for particular set of conditions, although in few tests we made the same measurement twice or more times with the same proband. Our focus here was to demonstrate the differences between the various PPEs so we preferred to measure 1 PPE with more persons to measuring 1 person several times under same conditions. However, as the lines L268-L275 (in the previous version) may have been confusing describing the data which were not used (all important information are covered by L202-208), I deleted the former article L268-L275 and decreased the number of tests included described on L471.

Table 4: Sorry that I missed it before, but if ReT of the 2 first clothing sets (TF & OPCH) with merino sweatshirt (b) was not measured, then where you got im values?

We deleted the confusing im numbers for TF/OPCH variant b.

In reality, for impermeable overalls TF and OPCH the variant b with merino sweatshirt was measured only for TF suit, but the influence of the sweatshirt was almost negligible (not for It!) within the error of measurement so the ReT value for variant a was used for the im calculation (in previous version of the paper). Under this experience, the ReT for OPCH variant b was not measured but the im was calculated based on ReT measured for variant a. And the im values stayed in the table from previous working calculations. But, to be correct, as the ReT for TF/OPCH were not finally calculated/measured, the im values for variant b were deleted. Thanks for noticing.

L613, … on average at 38.2 °C in average …: delete “in average” at the end of the line. (By the way, Thermal Work Limit (TWL) suggests using limiting core temperature of 38.2 °C, e.g. see Miller and Bates 2007).

Comment added on L732-736.

L723: See the comment on L613 – your work seems also supporting TWL of 38.2 °C, while your warning in L764-766 may support 38.0 °C to be on safe side (as is ILO/WHO recommendation).

Comment and references added on L763-773.