Sustainable Construction of Earth Dams: Use of Heterogeneous Material from the Dam Site

Round 1

Reviewer 1 Report

1) The phenomenon of overtopping should be explained and discussed in depth.

2) Juktan Dam, Torblaa and Rikartsen = Not correct way of writing.

3) What factors influence the permeability values present in a moraine material?

4) Give real examples of earth-filled dam with different core shapes. In which countries and their functions.

5) In what range does the permeability coefficient of the heterogeneous material vary? AND Why?

6) In which studies do we consider movement restrictions?

7) Can large dispersion of permeabilities affect the dam construction and its longevity?

8) As commented = Where?

9) Dont write two lines as ONE paragraph. Check and merge.

10) Remove GRID lines in all the graphs.

11) Dont write methodology in the results and discussion.

12) Are all the equations your results? OR materials and methods?

13) Increase the size of numbers in the X and Y axis.

14) Multiple figures = PLACE ONE BELOW other and enlarge. DONT PLACE SIDE BY SIDE.

15) Unit in Y axis = spacing mistakes

16) Units are not the same in the text, tables and figures = WHY?

17) Page 21 = Too many short and small paragraphs = LOOKS SO UGLY. Can the authors KINDLY MERGE paragraphs.

18) Equation 7 = Is it your result or your discussion?

19) Write conclusions in <150 words, one PARAGRAPH.

Move all text to the section before the conclusions.

 

 

Author Response

We are very grateful to the reviewer for the time spent on the paper and their valuable comments to improve it. All comments are very appropriate and the indicated modifications have been made.

The detailed answer is attached in a text document.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper presented a probabilistic analysis for the impermeable core of earth dams by considering the uncertainties and spatial variation in the permeability coefficient. Two output results, flow rate and maximum hydraulic gradient, were investigated in a probabilistic framework. Extensive simulations were performed accounting for different mean and variance values of the core permeability within two configurations: isotropic and anisotropic. Some design charts were proposed which permit to quickly estimate the failure probability of an earth dam once the required input parameters have been quantified. Overall, the manuscript was well written, the topic is interesting and the provided design charts could be interesting for practical engineering. Therefore, the paper could be considered for publication after a revision of considering the following remarks:

1. The expression ‘Montecarlo’ should be verified since it is generally written as ‘Monte Carlo Simulation/Method/Analysis’
2. This paper studied only the core part of a zoned earth dam. Then, how to link the obtained results of this simplified model (only core) to the safety assessment of the original dam in which the upstream and downstream shoulders are considered. The estimated flow rate and the maximum gradient within the core are representative for the entire part of a real zoned earth dam?
3. Section 4.2: one lift has a dimension of 4m*0.33m. Please justify the selection of such a dimension for the lift. For this reviewer, the dimension should be determined with respect to the autocorrelation distance of permeability coefficient in respectively horizontal and vertical direction.
4. This paper proposed a new and simplified way to consider the spatial variation of a material property by dividing the physical domain to a number of blocks. However, the correlation between the values assigned to the different lifts were not mentioned. Have the authors considered this issue, called as autocorrelation in space which is common for natural soils and should be found also in the core if it was made by the soils/sands from one site?
5. P7: please explain why saturated conditions were assumed? What will be the effects on the flow rate and maximum gradient by considering unsaturated flows?
6. P8: ‘and unbound functions such as the beta distribution function were not considered’. This reviewer cannot agree with such a statement. The Beta distribution is defined by two shape parameters in the interval [0, 1]. It can be transformed to any intervals, but this distribution has clearly two bounds.
7. P9: ‘have been fixed and imposing’. Imposed
8. P14: why sigmoid type was selected to fit the distribution? Any justification?
9. P20: ‘the flow rates through the core are always higher in the calculations made considering the homogeneous core permeability than in the calculations that were made considering the permeability of the core as heterogeneous.’ In which figures, the results of homogeneous core permeability were compared with the ones of heterogeneous cases?
10. The proposed design charts were obtained by considering a specific dam case which means the geometry is the one of Fig. 9, the water level was fixed at the dam top and the Kh/Kv ratio was set as 10. Then, how can these charts to be used for general cases (e.g. a different slope angle, a lower water level or/and a different Kh/Kv ratio)?
11. Section 5.3: the employed method is an approximative solution to estimate the failure probability. The accuracy in terms of the provided failure probability was not validated. In fact, a small failure probability (e.g. 0.0001), corresponding to the area of the small left/right tail of the distribution, could be very sensitive to the parameters used to describe the distribution.
12. P23: please explain how the limit value of 55.35 m2 was obtained? What was the value of k in the equation (Q/k) for anisotropic cases?
13. Figures 20-23: The points are related to the 16 cases in Table 2. But how these curves were obtained which normally needs a number of CoV and the corresponding failure probability? Using the fitted empirical distribution (Eq. (6))? But each equation in Table 3 is only for one CoV.
14. P30: ‘is characterized as a random variable with a lognormal distribution function’. Better to write as ‘a series of random variables’ or ‘a simplified random field’, since a single random variable can only describe a homogenous material.
15. Please have a look on the following references which deal with the relative topics to this paper, and add them into this manuscript if possible.
    • Mouyeaux A, Carvajal C, Bressolette P, et al. Probabilistic analysis of pore water pressures of an earth dam using a random finite element approach based on field data[J]. Engineering Geology, 2019, 259: 105190.
    • Guo X, Dias D, Pan Q. Probabilistic stability analysis of an embankment dam considering soil spatial variability[J]. Computers and Geotechnics, 2019, 113: 103093.
    • Guo X, Dias D, Carvajal C, et al. A comparative study of different reliability methods for high dimensional stochastic problems related to earth dam stability analyses[J]. Engineering Structures, 2019, 188: 591-602.

 

In summary, a decision of ‘Major Revision’ is recommended.

Author Response

We sincerely appreciate the comments of the reviewer to allow us to improve the objective and progress that is achieved with this research.

The detailed answer is attached in a text document.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have carried out all the revisions. I don't have any further comments.

Author Response

Thanks!

Reviewer 2 Report

This reviewer appreciates the efforts made by the authors to revise the manuscript. All the questions have been answered by the authors and their responses are adequate, except three remarks of my previous review report for which I have still some doubts.

1. Point 4: For one simulation, the different values assigned to the different lifts follow a Lognormal distribution. I agree with this statement. However, it doesn’t mean that the assigned values have correlation with each other. The autocorrelation in a random field describes the spatial correlation in space: normally, the values assigned to adjacent lifts should have a relatively strong correlation and such correlation is decreased with increasing the distance between two lifts. This needs an autocorrelation function with together the autocorrelation distances. According to section 4.5, the generated values from the Lognormal distribution are randomly assigned to the different lifts, so the authors probably have not considered the spatial correlation. This is one limitation of the proposed simplified approach compared to sophisticated random-field generation methods [1]. Please make clear about this point in the article.
2. Point 10: Ok for the response. I would suggest the authors to clearly express in the article that the proposed charts are suitable for the dams with similar configuration to the studied one, and new charts are necessary for significantly different cases.
3. Point 13: I am still unclear about how these curves were obtained. The fours points are from the performed Monte Carlo Analysis in combination with Eq. (6). Then, the curve is obtained by fitting to these four points? Which forms of equations were used for the fitting? Polynomials? Only four points are sufficient for a fitting? In addition, what it means ‘boundary conditions’? Also, ‘some point is’ should be replaced by ‘some points are’.

 

[1] Sudret B, Der Kiureghian A. Stochastic finite element methods and reliability: a state-of-the-art report[M]. Berkeley, CA: Department of Civil and Environmental Engineering, University of California, 2000.

Author Response

We especially thank this reviewer for the time taken to improve the article and the speed of the review process.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

The answers provided the authors are satisfactory. Now, the manuscript can be considered for publication after implementing the two following suggestions which are related to the last question of my previous review report:

1. I agree that it is not necessary to give details about how these curves were obtained. However, the name of the equation used for the fitting should be at least mentioned. Such information is important for readers to understand the adopted analysis procedure and could be useful for the researches who would like to repeat the study or compare with their own projects.
2. It is hard to judge the quality of a fitting by just looking at the number of points. It is better to present a statistical indicator which can quantify the fitting quality for the curves. The indicator could be coefficient of determination or root-mean-square error etc..

Author Response

We thank to the reviewer for the time taken to improve the article and the speed of the review process.

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The m/s addresses a fundamental and important aspect of dam construction and their materials. I just have a few technical comments:

1) The specific objectives and the conclusions should be aligned.

2) For lower permeability, generate localized zones - Provide good examples and case studies from different dams.

3) Classify the dams according to the types, functions, origin, life and discuss the results.

4) Compare the performance of dams/earth dams in Canada, Norway, Sweden, Finland, New Zealand, the former Soviet Union and France, its functions, etc in the form of a table. Give all dimensions required.

5) How did downstream phase by varying the effective length affect the result of this work? Give good illustrations.

6) What is the feature of compacted lifts ? Where can we apply them?

7) State the limitations and assumptions of Finite Difference Method.

8) Which previous studies have used Heterogeneous Material for the Dam Core? Please compare all these studies and their results with this work. I am expecting another 2-3 pages more of technical discussion + extra revision.

9) Compare the Various lognormal distribution functions in the form of a table and show all the important previous results.

10) X axis graphs numbers are CLUSTERED. Leave space.

11) Check reference formatting MANUALLY.

12) How did higher flow rate affect the overall conclusion of the study? 

13) What are the salient features of obtaining maximum gradients for a DAM construction project. Give case studies and examples.

14) WHich other studies have simulated the construction of these cores using heterogeneous materials? Give comparative analysis of the results.

15) What are the advantages of anisotropic permeability distribution.

16) X and Y axis of all graphs = Black colour.

17) All the figures are BLURRED + we cannot read anything when printed.

18) IMPROVE clarity of all the images.

19) Check units.

20) Check references.