the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Poro-perm relations of Mesozoic carbonates and fault breccia, Araxos Promontory, NW Greece
Abstract. Aiming at assessing the porosity and permeability properties, we present the results of microstructural and laboratory measurements of density, porosity, VP, VS, and electrical resistivity performed in dry and in saturated conditions on 54 blocks of Mesozoic carbonate host rocks and fault breccias. Host rocks consist of carbonate mudstones, wackestones, packstones, and sedimentary breccias pertaining to the Senonian and Vigla formations. These rocks show average density values, low values of porosity, and medium-to-high P- and S-wave velocities. Fault breccias derive from high-angle extensional and strike-slip fault zones, and are characterized by a wider range of density, porosity values up to 5–10 times higher than host rock, and low ultrasonic velocities. Independently on lithology, the carbonate host rocks might include vugs due to selective dissolution. Differently, the fault breccia samples include microfractures. A slight textural anisotropy is documented in the carbonate host rocks, whereas a higher degree of anisotropy characterizes the fault breccias. Selected samples were also tested in pressure vessels with confining pressure up to 80 MPa, showing that transport properties along microcracks in fault breccias can significantly increase with increasing depth. In order to assess rock permeability and porosity-permeability relations, three different protocols are employed. Two of them are based on the Effective Medium Theory, so that permeability is computed by inverting ultrasonic measurements and assuming an array of penny-shaped cracks embedded in an impermeable host matrix. Accordingly, the aspect ratio and crack width are obtained by the seismic measurements. Two end terms have been modelled by assuming all cracks isolated, and unconnected or all cracks connected into the network. Application of these two protocols shows a systematic variation of permeability with porosity, whereas the results of the third one, based on the digital image analysis outcomes, do not exhibit systematic variation. We interpret this behavior as due to the not-selective dissolution of the outcropping carbonates causing a wide range of measured fracture aperture values.
- Preprint
(2269 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Reviewer comment on egusphere-2023-2247', Anonymous Referee #1, 01 Dec 2023
This manuscript presents petrophysical data for limestones collected within or adjacent to a fault in Greece. There are several aspects of this manuscript that require attention, as explained in detail in my specific comments below. (i) There appears to be a lot of key information missing from the methods section. (ii) Graphs without an x-axis, and those that show data as a function of sample number, should be changed to "distance from the fault slip surface (meters)". As it stands, I think there is plenty of scope for misreading or misinterpreting these graphs. (iii) Some of the data promised in the methods section are not shown in the manuscript. (iv) The authors state that their main goal is to provide porosity-permeability relationships for the host and fault rocks (see also the title of the manuscript); however, Figure 19 only shows estimated permeability values for the host rocks, which show no trend. The authors should plot all their estimated permeability values on Figure 19 and discuss the relationship between porosity and permeability for both the host rocks and the fault rocks in more detail. (v) The discussion section requires a lot of work. The interpretation of the data is poorly and confusingly presented and explained, and the authors should better compare (and discuss) their data with those published previously for limestone. (vi) The limestones are very heterogenous, which calls into question the representativity of the data. Although I don't think this is a problem, the authors should spend more time describing, documenting, and discussing the heterogeneous nature of their samples. For the permeability tests, for example, the authors do not state in which orientation they cored their samples. (vii) There are numerous typos, spelling mistakes, and grammatical errors. The authors should carefully go through their manuscript and correct these mistakes. To conclude, there is quite a lot of data in this manuscript, and they're definitely of interest to the community. However, I feel as though there is still quite a lot of work to do to elevate the manuscript sufficiently to warrant publication. My recommendation is to reject the manuscript now, but encourage the authors to revise and resubmit their manuscript when the necessary (those deemed important by the handling editor) corrections are complete.
Specific comments
L. 17:- I think the authors should state that the rocks are from Greece.
L. 24-32:- This part of the abstract explains more about what the authors did, rather than what they found. Can the authors describe the main results and conclusions of their study here?
L. 34-36:- Can the authors provide a reference for this statement?
Figure 1:- I think that the authors should provide a figure, or another panel in this figure, that shows the sampling locations for their materials.
L. 112:- Were the blocks oriented?
L. 116-120:- Can the authors provide, for all rock samples, their distances away from the fault slip surface? This information is surely very important for the interpretation of the petrophysical data?
L. 120:- Given that the rocks are rather heterogenous, I think it's very important in this kind of study to provide some images of the rock blocks or outcrops. Perhaps the authors could provide representative photographs here, and then provide a photograph of each of the collected samples in a supplementary file?
L. 120:- How many blocks were collected for each type?
L. 124:- How many rock slabs were prepared for each of the five categories of rock outlined in L. 116-120?
L. 126:- The authors should explain the orientation of their blocks, and how they were cut, in the previous subsection.
L. 126:- How were the slabs "manually smoothed"?
L. 142:- Given that some of the fragments within the fault rocks will be angular, I think it would also be interesting to provide the aspect ratio or sphericity of the hand-drawn grains or fragments. The authors may find that the angularity of the grains or fragments increases with proximity to the fault, which would be interesting.
L. 143:- So D0(grains) is a measure of the clast/survivor grain number density?
Figure 2:- It's impossible to see the numbers in the top right of the figure. Do the authors need to show this? Perhaps they could also provide the data as a digital supplementary file?
L. 161:- To confirm, the authors did not perform this sieving analysis on their samples?
L. 169:- There are five rock categories, but only four thin sections were prepared. Which of the five rock categories were chosen for thin sectioning? What was the justification for choosing these four samples? In what orientation were the thin sections prepared?
L. 176:- A pixel in these images is how many microns? This is important to understand whether this analysis was able, or not, to capture the microporosity.
L. 177:- It should be stated that this provides the connected porosity. And, perhaps, that the method very likely underestimates the porosity of carbonates, rocks that often contain abundant microporosity.
Figure 3:- The samples are very heterogeneous. Can the authors comment on the representivity of their porosity analysis?
L. 197:- By "specimen", do the authors mean the saw-cut rock blocks? Can this be made clear?
L. 198:- So the water was not vacuumed into the void space? Since these some of these rocks are rather low porosity and low permeability, does this not mean that the connected porosities would be underestimated (and perhaps by a lot)?
L. 201:- The dry mass was determined after the saturation process? At what temperature? For how long? How were the authors sure that the rocks were completely dry?
L. 211:- And one of these directions is parallel to the bedding or slip surface?
L. 217:- The dynamic elastic moduli?
L. 236:- Why were these three samples chosen? How heterogenous were the cores prepared for the permeability measurements (given that the cores have a diameter of 40 mm)? Is it possible to provide photographs of the three samples (here or in a supplementary file)?
L. 358:- In which direction were the samples cored? Perpendicular to the bedding or slip surface? Were the two CFB samples cored in two different directions? This information is very important for the understanding and interpretation of these data.
L. 240:- The measurements were made on water-saturated samples or dry samples? What was the pore fluid used? This information is very important for the understanding and interpretation of these data.
L. 250:- What was the temperature of measurement?
L. 254:- If gas was used as the pore fluid, did the authors check for the Klinkenberg and Forchheimer corrections?
Figure 4:- Can the authors indicate the orientation of the blocks in the photographs?
L. 275:- Why not also refer to Figure 4a and 4c here and discuss the grain/fragment size analysis? The authors stated in their methods section that quantitative grain/fragment and void space analysis was performed. Where are these data for the host rock samples (samples HR-AFZ and HR-PFZ)? Why not present and discuss these data here?
L. 282:- Where are the quantitative grain/fragment and void space analysis data for the FHR-FZ samples? Why not present and discuss these data here?
L. 287:- Where are the quantitative grain/fragment and void space analysis data for the CFB-FZ samples? Why not present and discuss these data here?
L. 292:- Porosity is in percent? Where are the quantitative grain/fragment and void space analysis data for the FCFB-FZ samples? Why not present and discuss these data here?
L. 305:- Can the authors better explain this manmade fracture?
Figures 6, 7, 8, 9, and 10 (and Figure 13):- The x-axis labels on panel (b) in these figures is missing. The data in panel (a) in these figures are plotted as histograms, and the data are organized by sample number. Although the sample groupings are ordered according to their relative distance (these distances are still unknown, see my comment above) from the main slip surface, is it true, for example, that AR17 is closer to the slip surface than AR145? If this is not the case, then these graphs, and the dashed lines, could be considered misleading. Can the authors not plot these data as a function of distance (in meters) from the main slip surface?
L. 315:- Elastic wave velocities were measured in three orthogonal directions on each block. The data shown in Figures 8 and 9 are the averages of these measurements? If so, I think this should be clearly stated (i.e. that the authors are averaging measurements taken in different directions).
L. 331 and elsewhere:- Why "textural anisotropy"? Are the authors not describing an "elastic wave anisotropy" here? Further, since anisotropy refers to a property, can texture be anisotropic? I would rather say that the texture was heterogeneous or that it showed a preferred orientation.
L. 338:- "There is not a clear directivity in the textural anisotropy." Is anisotropy not defined as a "property with a directional dependence"? I find this sentence confusing. Do the authors mean that the samples are homogeneous and are not characterized by an anisotropy?
Figure 10:- It's not so easy to compare the various data with the data plotted in this way. Is there not a better way to present these data?
Figure 10:- The caption refers to (a) and (b), but there is only one panel.
L. 341:- This is interpretation and should be moved to the discussion section.
L. 344:- Can the authors refer back to photographs of the samples that show these features (bedding and textural heterogeneity associated with the slip surface)?
L. 349:- Can the authors refer to a photograph that provides evidence for these fractures?
L. 354:- Can the authors refer to a photograph that provides evidence for these fractures?
L. 356:- Can the authors refer to a photograph that provides evidence for these fractures?
Figure 11 (and Figure 12):- As for some of the panels in the previous figures, there is no x-axis on the panels in this figure. "Distance from slip surface (m)" would be ideal here, in my opinion. I think it's misleading for the reader to provide XY plots without an x-axis, because the reader may naturally assume that these rocks were collected at equally-spaced distances from the slip surface. I think the authors should retrieve their GPS or location data (if they have them) for each of their sampling locations and plot these data as a function of distance from the slip surface.
Figure 11:- The authors also determined shear modulus. Why not show this here too?
L. 372:- The authors did not explain that they measured this property in both wet and dry conditions. What fluid was used? Distilled water? The authors should provide more details on this in their methods section.
L. 373:- The authors interpret that the fault rocks have a higher electrical resistivity value than the other rocks. However, looking at Figure 12, I would conclude that there is no trend. What are the errors on these data? Can the authors actually confidently conclude that the electrical resistivity value is higher in the fault rock samples?
L. 374:- The authors do not describe the influence of fluid-saturation on the data.
L. 380:- "hypothetical distance". I have an issue with the authors visualising data in this way. I think it's misleading, as described in my comments above. Can the authors not determine a distance from the slip surface for their various sampling locations?
L. 381:- Why not plot Young's modulus as a function of porosity (and color-code the symbols to represent the rock type)? In fact, it's typical to plot rock physical properties as a function of porosity. Why not do this for all of the properties measured (instead of plotting them as a function of "hypothetical distance")?
Figure 13:- It is stated in the figure caption that the shear modulus is shown by black bars. But, there are no black bars. What about Poisson's ratio?
L. 391:- This is interpretation and should be moved to the discussion section.
Figure 14 (and throughout):- Can the authors use the SI units for permeability?
L. 420:- Did the authors use the wet dynamic moduli?
L. 437:- The authors used their dry data? If so, is this not a problem if the permeability measurements were performed using water as the pore fluid (information missing from the methods section)?
L. 438:- This is not well explained in the methods section. Did the authors use some threshold value of aspect ratio to differentiate pores and cracks?
L. 448:- So the authors only included the crack porosity here? How did the authors differentiate pores and cracks, using the aspect ratio?
L. 458:- Why did the authors assume a crack width of 1 micron? Is this in line with the data? Are cracks with an aperture of 1 micron resolvable on the images?
L. 460:- What are the reference values for the Young's modulus and Poisson's ratio? The authors should also cite the source of these data.
Figure 16:- From what I can gather, D0(grains) is a measure of the number density of clast/survivor grains. It would seem logical that this is related to the amount of matrix. What are the authors trying to show with this figure? What is the purpose of the empirical relationship shown on the figure? D0(grains) as a function of distance from the slip surface would be much more informative.
Figure 16:- Is there actually a relationship between proximity to the slip surface and the number density of clast/survivor grains? There are host rock samples with high grain densities, and fault rock samples with low grain densities.
L. 498-501:- I don't follow the argument here. Can the authors better explain their interpretation of these data?
Figure 17:- Why are all the P-wave velocities below 1 km/s?
L. 508:- What is the "best-fit" line? This is an empirical fit to a larger dataset?
L. 514:- Are molds observed in the microstructure?
L. 517:- Are vugs observed in the microstructure? Are these microstructural interpretations in line with the microstructural observations?
L. 519:- What does the best-fit line represent? If it's an empirical fit to a larger dataset, what is the significance of the data aligning with this fit?
L. 521:- Are the more limestone data that can be included in Figure 17 for comparison?
L. 532:- Is this text supposed to be in bold? It's part of the caption, or is it part of the text below?
Tables 1 and 2:- Why are these data presented in the discussion section? Would it not be more appropriate to present and describe these data in the results section?
L. 554:- What is the significance of "pores are more uniformly distributed in higher porosity carbonates"? Why would the pores be more uniformly distributed in the fault rocks? I think it would be interesting to comment on how the shape of the void space is changing (or not) as a function of distance from the fault slip surface. The authors have a lot of statistical information about their porosity from their ImageJ analysis: why not discuss this data here?
L. 560:- I think the authors should clearly state that these are permeability estimations.
Figure 19:- The authors suggest at the start of their discussion section that they would compare their data with those published previously. I think it would be interesting to compare their porosity-permeability data with other published data (data other than the limestones from Italy). Are the samples measured here more or less permeable than other limestones?
Figure 19:- The authors only show permeability estimations for the host rocks. What about the estimations for the fault rocks? The authors state in their introduction that finding porosity-permeability relationships for the host and fault rocks is the aim of their study. Surely all of the permeability estimations from Figure 15 should be included in Figure 19?
L. 569-570:- Why does the presence of "stiff, sub-rounded pores" explain the lack of correlation between porosity and permeability?
L. 572:- Is the increase really linear (it's a log-linear scale)?
L. 574:- The papers by Benson et al. (2006) and Nasseri et al. (2007) do not provide data for limestones. I think the authors should state the rock types measured in these studies, rather than stating "cohesive rocks". What can the authors say about porosity-permeability relationships in other limestones? Are they also linear? It's odd that the authors do not compare their data with other trends found for limestone.
L. 577:- How did the authors decide on these "qualitative" trends? Further, since the x-axis is sample number (see my comments above), what does this trend mean?
L. 578:- The authors suggest in their introduction that the data presented in their manuscript would be useful for reservoir assessment or exploration. I think it would be interesting, because of this, to outline the implications of their data for reservoir assessment or exploration. Furthermore, the authors should also outline the limitations of their data for this purpose. In other words, that any laboratory data would require upscaling before it can be considered relevant for the reservoir scale.
Citation: https://doi.org/10.5194/egusphere-2023-2247-RC1 -
RC2: 'Comment on egusphere-2023-2247', Anonymous Referee #2, 22 Dec 2023
Review of the manuscript #egusphere-2023-2247 entitled "Poro-perm relations of Mesozoic carbonates and fault breccia, Araxos Promontory, NW Greece" by Vinciguerra et al.
The manuscript by Vinciguerra et al. deals with the experimental characterisation of petrophysical properties of carbonates and fault breccia. In addition, micromechanical models based on crack population are used to infer permeability of the rocks by inverting ultrasonic measurements. Analysis of the evolution of the properties allow the authors to draw some conclusions about the fracture aperture distribution.
The manuscript is of potential interest to EGUsphere readers but several issues must be addressed before a revised version may be accepted for publication.
- First of all, numerous typos, mistakes and other grammatical errors scatter the text. The authors should meticulously go through the text and correct this point.
- The description of the geological setting (l. 75-106) is too long and should be shortened to focus on the points relevant to the objectives of the study.
- The description of the 2D microstructural analysis protocol (l. 124-161) is far too long and should be shortened to focus on the key question: why is measuring a fractal dimension relevant to the aims of the study? Especially measuring a 2D fractal dimension whereas porosity and permeability are 3D properties.
- l. 200: The authors give details on how the various masses are measured but not on how the volume of the sample is evaluated. They talk about the water displacement method without further details on the resolution of the method. They could have used the triple weighing method, which allows porosity to be determined without needing to know volume. The dry mass is evaluated in natural conditions without drying process. How do the authors guarantee that the sample is dry and contains no adsorbed residual water? Why not vacuum-dry the sample? Since some of the samples have a low porosity, this may be a crucial point the authors should address.
- l. 209-220: The authors infer elastic parameters from ultrasonic measurements. They should be aware that these parameters are dynamic ones that may be different from the static ones derived from stress-strain curves. The authors should address this point by comparing both sets of values to convince the reader that they are equivalent, or remove this reference to static parameters.
- l. 240-250: The authors measure permeability to gas without taking into account any possible correction due to non-laminar fluid flow conditions (Forchheimer and/or Klinkenberg). They also use two methods (constant flow and pulse decay) without showing how they compare. This is a crucial point because they make two mistakes: (1) the constant flow method at a fixed pressure gradient leads to an apparent permeability which may be different from the true (corrected) one, (2) using the pulse decay method means that the pressure gradient is continuously decreasing thus leading via Eq. (11) to a varying apparent permeability. Without introducing the required correction, the time decay parameter will be changing with time thus leading to a wrong permeability. The authors have to clearly show that the measured permeabilities are correct, since they are then used to calibrate the estimated values inferred from UPV.
- l. 280-290: Same question as before : how can fractal dimension measurement help to achieve the goals of the study? Apart from the numbers, what can we infer from these measurements?
- Figs. 6, 7, 8, 9, 10, 13, 15: All these plots use an x-axis without units. It causes a lot of confusion in the discussion. Maybe a distance would be more appropriate. This would also allow the authors to better explain how the various linear trends were obtained. A second point concerns the type of visualization used in these figures. Why histograms and not simply plots with symbols? This would also allow to put error bars on the data and give an idea of the accuracy of measurements.
- Figs. 8 and 9: For velocities, what is the bias introduced by averaging velocity on each block without taking into account the anisotropy? As shown in Fig. 10 the anisotropy is high for some samples. In fact Figs. 8 and 9 should be either split in four figures with the minimum and maximum velocities or replaced by usual plots with error bars related to the max-min velocity difference.
- Fig. 10: This is the limit of the histogram representation: empty and full bars are indistinguishable. Another representation is required.
- Fig. 12: what makes dry samples conductive? The fact that they are not really dry? The ratio wet/dry is lower than 10 which suggests that the drying method used in the paper is not efficient. Some comments are needed. Not all measurements have error bars. Why? Only one measurement or is the error lower than the size of the symbol? Putting a trend curve is rather misleading since there is no x-axis unit. Should be removed.
- l. 182: it is not possible to infer from Fig. 11 (no x-axis) or Fig. 13 (not legible) that dynamic modulus and porosity are correlated. To support this idea, Fig.13 should be replaced by a plot relating these two quantities.
- Section 4.2: following previous remark on apparent permeability, data shown in Fig. 14 may be questionable as they result from a mix of two methods without any correction. Either these corrections are negligible but the authors have to support this assumption, or the authors have to use corrected values.
- l. 455-460: In the modelling section, permeability is estimated from velocity measurements. In WF1, aspect ratio and crack width required to calculate permeability are directly derived from the UVP measurements. In WF2, a constant crack aperture of 1 µm is assumed. How do these two values compare? The crack width results derived from UVP have to be shown and compared to the 1 µm estimate in order to convince the reader that there is no "black box" effect in the derivation. Is the 1 µm value consistent with the image analysis presented before?
- l. 460: How are the reference elastic modulus and Poisson's ratio determined? From literature? What justifies the use of generic values for this particular study? From dedicated stress tests by using the slope of stress-strain curves as suggested previously by the authors? More details are needed to clarify this point.
- l. 460-466: The authors use the percolation factor as an adjusting variable but recognize that it is very complex to measure. The authors' approach is perhaps not the simplest and should be revisited. Since all samples show a rather high permeability, they could simply assume that for all them we are well above the percolation threshold with a well-connected flowpath and thus simply assume that f is equal to 1. This would allow the modelling to rely entirely on microstructural parameters that can be estimated either by UPV, image analysis or other non-destructive tests. The authors should at least compare both approaches to convince the reader.
- Fig. 15: There are only three control points and it is very difficult to estimate the accuracy of the method, as the type of diagram used – see comments above – does not help the reader. Since this figure can be seen as the main objective of the paper and a strong support to the approach, it needs to be reworked. Another important point is that more permeability measurements are needed, especially in zones where this property varies a lot to support the relevance of the approach.
- Fig. 17: Porosity is normalized means that there is no unit (%).
To summarize, this paper presents a collection of new data that makes it of potential interest to EGUsphere readers. However the main objective of the article, i.e. the presentation of permeability-porosity relationships, is not achieved. This is certainly due to the heterogeneous nature of the rocks studied, but also to the proposed approach, using a model with a fitting variable that is not the best, and to the very low number of calibration points. It is highly recommended to reconsider the model in the way proposed above, and to have more real permeability measurements to calibrate the model before deciding on the best approach. I cannot recommend publication of the manuscript in its present state but encourage the authors to do the necessary corrections and complementary work before submitting a new version of this potentially interesting paper.
Citation: https://doi.org/10.5194/egusphere-2023-2247-RC2
Status: closed
-
RC1: 'Reviewer comment on egusphere-2023-2247', Anonymous Referee #1, 01 Dec 2023
This manuscript presents petrophysical data for limestones collected within or adjacent to a fault in Greece. There are several aspects of this manuscript that require attention, as explained in detail in my specific comments below. (i) There appears to be a lot of key information missing from the methods section. (ii) Graphs without an x-axis, and those that show data as a function of sample number, should be changed to "distance from the fault slip surface (meters)". As it stands, I think there is plenty of scope for misreading or misinterpreting these graphs. (iii) Some of the data promised in the methods section are not shown in the manuscript. (iv) The authors state that their main goal is to provide porosity-permeability relationships for the host and fault rocks (see also the title of the manuscript); however, Figure 19 only shows estimated permeability values for the host rocks, which show no trend. The authors should plot all their estimated permeability values on Figure 19 and discuss the relationship between porosity and permeability for both the host rocks and the fault rocks in more detail. (v) The discussion section requires a lot of work. The interpretation of the data is poorly and confusingly presented and explained, and the authors should better compare (and discuss) their data with those published previously for limestone. (vi) The limestones are very heterogenous, which calls into question the representativity of the data. Although I don't think this is a problem, the authors should spend more time describing, documenting, and discussing the heterogeneous nature of their samples. For the permeability tests, for example, the authors do not state in which orientation they cored their samples. (vii) There are numerous typos, spelling mistakes, and grammatical errors. The authors should carefully go through their manuscript and correct these mistakes. To conclude, there is quite a lot of data in this manuscript, and they're definitely of interest to the community. However, I feel as though there is still quite a lot of work to do to elevate the manuscript sufficiently to warrant publication. My recommendation is to reject the manuscript now, but encourage the authors to revise and resubmit their manuscript when the necessary (those deemed important by the handling editor) corrections are complete.
Specific comments
L. 17:- I think the authors should state that the rocks are from Greece.
L. 24-32:- This part of the abstract explains more about what the authors did, rather than what they found. Can the authors describe the main results and conclusions of their study here?
L. 34-36:- Can the authors provide a reference for this statement?
Figure 1:- I think that the authors should provide a figure, or another panel in this figure, that shows the sampling locations for their materials.
L. 112:- Were the blocks oriented?
L. 116-120:- Can the authors provide, for all rock samples, their distances away from the fault slip surface? This information is surely very important for the interpretation of the petrophysical data?
L. 120:- Given that the rocks are rather heterogenous, I think it's very important in this kind of study to provide some images of the rock blocks or outcrops. Perhaps the authors could provide representative photographs here, and then provide a photograph of each of the collected samples in a supplementary file?
L. 120:- How many blocks were collected for each type?
L. 124:- How many rock slabs were prepared for each of the five categories of rock outlined in L. 116-120?
L. 126:- The authors should explain the orientation of their blocks, and how they were cut, in the previous subsection.
L. 126:- How were the slabs "manually smoothed"?
L. 142:- Given that some of the fragments within the fault rocks will be angular, I think it would also be interesting to provide the aspect ratio or sphericity of the hand-drawn grains or fragments. The authors may find that the angularity of the grains or fragments increases with proximity to the fault, which would be interesting.
L. 143:- So D0(grains) is a measure of the clast/survivor grain number density?
Figure 2:- It's impossible to see the numbers in the top right of the figure. Do the authors need to show this? Perhaps they could also provide the data as a digital supplementary file?
L. 161:- To confirm, the authors did not perform this sieving analysis on their samples?
L. 169:- There are five rock categories, but only four thin sections were prepared. Which of the five rock categories were chosen for thin sectioning? What was the justification for choosing these four samples? In what orientation were the thin sections prepared?
L. 176:- A pixel in these images is how many microns? This is important to understand whether this analysis was able, or not, to capture the microporosity.
L. 177:- It should be stated that this provides the connected porosity. And, perhaps, that the method very likely underestimates the porosity of carbonates, rocks that often contain abundant microporosity.
Figure 3:- The samples are very heterogeneous. Can the authors comment on the representivity of their porosity analysis?
L. 197:- By "specimen", do the authors mean the saw-cut rock blocks? Can this be made clear?
L. 198:- So the water was not vacuumed into the void space? Since these some of these rocks are rather low porosity and low permeability, does this not mean that the connected porosities would be underestimated (and perhaps by a lot)?
L. 201:- The dry mass was determined after the saturation process? At what temperature? For how long? How were the authors sure that the rocks were completely dry?
L. 211:- And one of these directions is parallel to the bedding or slip surface?
L. 217:- The dynamic elastic moduli?
L. 236:- Why were these three samples chosen? How heterogenous were the cores prepared for the permeability measurements (given that the cores have a diameter of 40 mm)? Is it possible to provide photographs of the three samples (here or in a supplementary file)?
L. 358:- In which direction were the samples cored? Perpendicular to the bedding or slip surface? Were the two CFB samples cored in two different directions? This information is very important for the understanding and interpretation of these data.
L. 240:- The measurements were made on water-saturated samples or dry samples? What was the pore fluid used? This information is very important for the understanding and interpretation of these data.
L. 250:- What was the temperature of measurement?
L. 254:- If gas was used as the pore fluid, did the authors check for the Klinkenberg and Forchheimer corrections?
Figure 4:- Can the authors indicate the orientation of the blocks in the photographs?
L. 275:- Why not also refer to Figure 4a and 4c here and discuss the grain/fragment size analysis? The authors stated in their methods section that quantitative grain/fragment and void space analysis was performed. Where are these data for the host rock samples (samples HR-AFZ and HR-PFZ)? Why not present and discuss these data here?
L. 282:- Where are the quantitative grain/fragment and void space analysis data for the FHR-FZ samples? Why not present and discuss these data here?
L. 287:- Where are the quantitative grain/fragment and void space analysis data for the CFB-FZ samples? Why not present and discuss these data here?
L. 292:- Porosity is in percent? Where are the quantitative grain/fragment and void space analysis data for the FCFB-FZ samples? Why not present and discuss these data here?
L. 305:- Can the authors better explain this manmade fracture?
Figures 6, 7, 8, 9, and 10 (and Figure 13):- The x-axis labels on panel (b) in these figures is missing. The data in panel (a) in these figures are plotted as histograms, and the data are organized by sample number. Although the sample groupings are ordered according to their relative distance (these distances are still unknown, see my comment above) from the main slip surface, is it true, for example, that AR17 is closer to the slip surface than AR145? If this is not the case, then these graphs, and the dashed lines, could be considered misleading. Can the authors not plot these data as a function of distance (in meters) from the main slip surface?
L. 315:- Elastic wave velocities were measured in three orthogonal directions on each block. The data shown in Figures 8 and 9 are the averages of these measurements? If so, I think this should be clearly stated (i.e. that the authors are averaging measurements taken in different directions).
L. 331 and elsewhere:- Why "textural anisotropy"? Are the authors not describing an "elastic wave anisotropy" here? Further, since anisotropy refers to a property, can texture be anisotropic? I would rather say that the texture was heterogeneous or that it showed a preferred orientation.
L. 338:- "There is not a clear directivity in the textural anisotropy." Is anisotropy not defined as a "property with a directional dependence"? I find this sentence confusing. Do the authors mean that the samples are homogeneous and are not characterized by an anisotropy?
Figure 10:- It's not so easy to compare the various data with the data plotted in this way. Is there not a better way to present these data?
Figure 10:- The caption refers to (a) and (b), but there is only one panel.
L. 341:- This is interpretation and should be moved to the discussion section.
L. 344:- Can the authors refer back to photographs of the samples that show these features (bedding and textural heterogeneity associated with the slip surface)?
L. 349:- Can the authors refer to a photograph that provides evidence for these fractures?
L. 354:- Can the authors refer to a photograph that provides evidence for these fractures?
L. 356:- Can the authors refer to a photograph that provides evidence for these fractures?
Figure 11 (and Figure 12):- As for some of the panels in the previous figures, there is no x-axis on the panels in this figure. "Distance from slip surface (m)" would be ideal here, in my opinion. I think it's misleading for the reader to provide XY plots without an x-axis, because the reader may naturally assume that these rocks were collected at equally-spaced distances from the slip surface. I think the authors should retrieve their GPS or location data (if they have them) for each of their sampling locations and plot these data as a function of distance from the slip surface.
Figure 11:- The authors also determined shear modulus. Why not show this here too?
L. 372:- The authors did not explain that they measured this property in both wet and dry conditions. What fluid was used? Distilled water? The authors should provide more details on this in their methods section.
L. 373:- The authors interpret that the fault rocks have a higher electrical resistivity value than the other rocks. However, looking at Figure 12, I would conclude that there is no trend. What are the errors on these data? Can the authors actually confidently conclude that the electrical resistivity value is higher in the fault rock samples?
L. 374:- The authors do not describe the influence of fluid-saturation on the data.
L. 380:- "hypothetical distance". I have an issue with the authors visualising data in this way. I think it's misleading, as described in my comments above. Can the authors not determine a distance from the slip surface for their various sampling locations?
L. 381:- Why not plot Young's modulus as a function of porosity (and color-code the symbols to represent the rock type)? In fact, it's typical to plot rock physical properties as a function of porosity. Why not do this for all of the properties measured (instead of plotting them as a function of "hypothetical distance")?
Figure 13:- It is stated in the figure caption that the shear modulus is shown by black bars. But, there are no black bars. What about Poisson's ratio?
L. 391:- This is interpretation and should be moved to the discussion section.
Figure 14 (and throughout):- Can the authors use the SI units for permeability?
L. 420:- Did the authors use the wet dynamic moduli?
L. 437:- The authors used their dry data? If so, is this not a problem if the permeability measurements were performed using water as the pore fluid (information missing from the methods section)?
L. 438:- This is not well explained in the methods section. Did the authors use some threshold value of aspect ratio to differentiate pores and cracks?
L. 448:- So the authors only included the crack porosity here? How did the authors differentiate pores and cracks, using the aspect ratio?
L. 458:- Why did the authors assume a crack width of 1 micron? Is this in line with the data? Are cracks with an aperture of 1 micron resolvable on the images?
L. 460:- What are the reference values for the Young's modulus and Poisson's ratio? The authors should also cite the source of these data.
Figure 16:- From what I can gather, D0(grains) is a measure of the number density of clast/survivor grains. It would seem logical that this is related to the amount of matrix. What are the authors trying to show with this figure? What is the purpose of the empirical relationship shown on the figure? D0(grains) as a function of distance from the slip surface would be much more informative.
Figure 16:- Is there actually a relationship between proximity to the slip surface and the number density of clast/survivor grains? There are host rock samples with high grain densities, and fault rock samples with low grain densities.
L. 498-501:- I don't follow the argument here. Can the authors better explain their interpretation of these data?
Figure 17:- Why are all the P-wave velocities below 1 km/s?
L. 508:- What is the "best-fit" line? This is an empirical fit to a larger dataset?
L. 514:- Are molds observed in the microstructure?
L. 517:- Are vugs observed in the microstructure? Are these microstructural interpretations in line with the microstructural observations?
L. 519:- What does the best-fit line represent? If it's an empirical fit to a larger dataset, what is the significance of the data aligning with this fit?
L. 521:- Are the more limestone data that can be included in Figure 17 for comparison?
L. 532:- Is this text supposed to be in bold? It's part of the caption, or is it part of the text below?
Tables 1 and 2:- Why are these data presented in the discussion section? Would it not be more appropriate to present and describe these data in the results section?
L. 554:- What is the significance of "pores are more uniformly distributed in higher porosity carbonates"? Why would the pores be more uniformly distributed in the fault rocks? I think it would be interesting to comment on how the shape of the void space is changing (or not) as a function of distance from the fault slip surface. The authors have a lot of statistical information about their porosity from their ImageJ analysis: why not discuss this data here?
L. 560:- I think the authors should clearly state that these are permeability estimations.
Figure 19:- The authors suggest at the start of their discussion section that they would compare their data with those published previously. I think it would be interesting to compare their porosity-permeability data with other published data (data other than the limestones from Italy). Are the samples measured here more or less permeable than other limestones?
Figure 19:- The authors only show permeability estimations for the host rocks. What about the estimations for the fault rocks? The authors state in their introduction that finding porosity-permeability relationships for the host and fault rocks is the aim of their study. Surely all of the permeability estimations from Figure 15 should be included in Figure 19?
L. 569-570:- Why does the presence of "stiff, sub-rounded pores" explain the lack of correlation between porosity and permeability?
L. 572:- Is the increase really linear (it's a log-linear scale)?
L. 574:- The papers by Benson et al. (2006) and Nasseri et al. (2007) do not provide data for limestones. I think the authors should state the rock types measured in these studies, rather than stating "cohesive rocks". What can the authors say about porosity-permeability relationships in other limestones? Are they also linear? It's odd that the authors do not compare their data with other trends found for limestone.
L. 577:- How did the authors decide on these "qualitative" trends? Further, since the x-axis is sample number (see my comments above), what does this trend mean?
L. 578:- The authors suggest in their introduction that the data presented in their manuscript would be useful for reservoir assessment or exploration. I think it would be interesting, because of this, to outline the implications of their data for reservoir assessment or exploration. Furthermore, the authors should also outline the limitations of their data for this purpose. In other words, that any laboratory data would require upscaling before it can be considered relevant for the reservoir scale.
Citation: https://doi.org/10.5194/egusphere-2023-2247-RC1 -
RC2: 'Comment on egusphere-2023-2247', Anonymous Referee #2, 22 Dec 2023
Review of the manuscript #egusphere-2023-2247 entitled "Poro-perm relations of Mesozoic carbonates and fault breccia, Araxos Promontory, NW Greece" by Vinciguerra et al.
The manuscript by Vinciguerra et al. deals with the experimental characterisation of petrophysical properties of carbonates and fault breccia. In addition, micromechanical models based on crack population are used to infer permeability of the rocks by inverting ultrasonic measurements. Analysis of the evolution of the properties allow the authors to draw some conclusions about the fracture aperture distribution.
The manuscript is of potential interest to EGUsphere readers but several issues must be addressed before a revised version may be accepted for publication.
- First of all, numerous typos, mistakes and other grammatical errors scatter the text. The authors should meticulously go through the text and correct this point.
- The description of the geological setting (l. 75-106) is too long and should be shortened to focus on the points relevant to the objectives of the study.
- The description of the 2D microstructural analysis protocol (l. 124-161) is far too long and should be shortened to focus on the key question: why is measuring a fractal dimension relevant to the aims of the study? Especially measuring a 2D fractal dimension whereas porosity and permeability are 3D properties.
- l. 200: The authors give details on how the various masses are measured but not on how the volume of the sample is evaluated. They talk about the water displacement method without further details on the resolution of the method. They could have used the triple weighing method, which allows porosity to be determined without needing to know volume. The dry mass is evaluated in natural conditions without drying process. How do the authors guarantee that the sample is dry and contains no adsorbed residual water? Why not vacuum-dry the sample? Since some of the samples have a low porosity, this may be a crucial point the authors should address.
- l. 209-220: The authors infer elastic parameters from ultrasonic measurements. They should be aware that these parameters are dynamic ones that may be different from the static ones derived from stress-strain curves. The authors should address this point by comparing both sets of values to convince the reader that they are equivalent, or remove this reference to static parameters.
- l. 240-250: The authors measure permeability to gas without taking into account any possible correction due to non-laminar fluid flow conditions (Forchheimer and/or Klinkenberg). They also use two methods (constant flow and pulse decay) without showing how they compare. This is a crucial point because they make two mistakes: (1) the constant flow method at a fixed pressure gradient leads to an apparent permeability which may be different from the true (corrected) one, (2) using the pulse decay method means that the pressure gradient is continuously decreasing thus leading via Eq. (11) to a varying apparent permeability. Without introducing the required correction, the time decay parameter will be changing with time thus leading to a wrong permeability. The authors have to clearly show that the measured permeabilities are correct, since they are then used to calibrate the estimated values inferred from UPV.
- l. 280-290: Same question as before : how can fractal dimension measurement help to achieve the goals of the study? Apart from the numbers, what can we infer from these measurements?
- Figs. 6, 7, 8, 9, 10, 13, 15: All these plots use an x-axis without units. It causes a lot of confusion in the discussion. Maybe a distance would be more appropriate. This would also allow the authors to better explain how the various linear trends were obtained. A second point concerns the type of visualization used in these figures. Why histograms and not simply plots with symbols? This would also allow to put error bars on the data and give an idea of the accuracy of measurements.
- Figs. 8 and 9: For velocities, what is the bias introduced by averaging velocity on each block without taking into account the anisotropy? As shown in Fig. 10 the anisotropy is high for some samples. In fact Figs. 8 and 9 should be either split in four figures with the minimum and maximum velocities or replaced by usual plots with error bars related to the max-min velocity difference.
- Fig. 10: This is the limit of the histogram representation: empty and full bars are indistinguishable. Another representation is required.
- Fig. 12: what makes dry samples conductive? The fact that they are not really dry? The ratio wet/dry is lower than 10 which suggests that the drying method used in the paper is not efficient. Some comments are needed. Not all measurements have error bars. Why? Only one measurement or is the error lower than the size of the symbol? Putting a trend curve is rather misleading since there is no x-axis unit. Should be removed.
- l. 182: it is not possible to infer from Fig. 11 (no x-axis) or Fig. 13 (not legible) that dynamic modulus and porosity are correlated. To support this idea, Fig.13 should be replaced by a plot relating these two quantities.
- Section 4.2: following previous remark on apparent permeability, data shown in Fig. 14 may be questionable as they result from a mix of two methods without any correction. Either these corrections are negligible but the authors have to support this assumption, or the authors have to use corrected values.
- l. 455-460: In the modelling section, permeability is estimated from velocity measurements. In WF1, aspect ratio and crack width required to calculate permeability are directly derived from the UVP measurements. In WF2, a constant crack aperture of 1 µm is assumed. How do these two values compare? The crack width results derived from UVP have to be shown and compared to the 1 µm estimate in order to convince the reader that there is no "black box" effect in the derivation. Is the 1 µm value consistent with the image analysis presented before?
- l. 460: How are the reference elastic modulus and Poisson's ratio determined? From literature? What justifies the use of generic values for this particular study? From dedicated stress tests by using the slope of stress-strain curves as suggested previously by the authors? More details are needed to clarify this point.
- l. 460-466: The authors use the percolation factor as an adjusting variable but recognize that it is very complex to measure. The authors' approach is perhaps not the simplest and should be revisited. Since all samples show a rather high permeability, they could simply assume that for all them we are well above the percolation threshold with a well-connected flowpath and thus simply assume that f is equal to 1. This would allow the modelling to rely entirely on microstructural parameters that can be estimated either by UPV, image analysis or other non-destructive tests. The authors should at least compare both approaches to convince the reader.
- Fig. 15: There are only three control points and it is very difficult to estimate the accuracy of the method, as the type of diagram used – see comments above – does not help the reader. Since this figure can be seen as the main objective of the paper and a strong support to the approach, it needs to be reworked. Another important point is that more permeability measurements are needed, especially in zones where this property varies a lot to support the relevance of the approach.
- Fig. 17: Porosity is normalized means that there is no unit (%).
To summarize, this paper presents a collection of new data that makes it of potential interest to EGUsphere readers. However the main objective of the article, i.e. the presentation of permeability-porosity relationships, is not achieved. This is certainly due to the heterogeneous nature of the rocks studied, but also to the proposed approach, using a model with a fitting variable that is not the best, and to the very low number of calibration points. It is highly recommended to reconsider the model in the way proposed above, and to have more real permeability measurements to calibrate the model before deciding on the best approach. I cannot recommend publication of the manuscript in its present state but encourage the authors to do the necessary corrections and complementary work before submitting a new version of this potentially interesting paper.
Citation: https://doi.org/10.5194/egusphere-2023-2247-RC2
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
197 | 58 | 16 | 271 | 17 | 15 |
- HTML: 197
- PDF: 58
- XML: 16
- Total: 271
- BibTeX: 17
- EndNote: 15
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1