Technical auditing of rock mechanics modelling and rock engineering design

Technical auditing of rock mechanics modelling and rock engineering design

International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886 Contents lists available at ScienceDirect International Journal of Rock ...

981KB Sizes 5 Downloads 82 Views

International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

Contents lists available at ScienceDirect

International Journal of Rock Mechanics & Mining Sciences journal homepage: www.elsevier.com/locate/ijrmms

Technical auditing of rock mechanics modelling and rock engineering design John A. Hudson a,n, Xia-Ting Feng b a b

Department of Earth Science and Engineering, Imperial College, London SW7 2AZ, UK State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan 430071, China

a r t i c l e in fo

abstract

Article history: Received 17 March 2008 Received in revised form 22 April 2010 Accepted 10 May 2010 Available online 26 June 2010

We present a procedure for technically auditing rock mechanics modelling and rock engineering design. As rock mechanics modelling becomes more sophisticated, with the ability to include more parameters and to be coupled with other disciplines such as hydrogeology, and as the requirements on rock engineering projects become ever more challenging, it is prudent, if not essential, to have some form of procedure for checking that the modelling and design are suitable within the limits of current knowledge. Accordingly, we present here an appropriate auditing procedure, and demonstrate its use with the examples of measurement of rock stress (as a soft audit) and modelling for the design of hydropower caverns in China (as a semi-hard audit), the auditing questions being tailored to the nature of the work being audited. The types of procedures demonstrated can be used from the initial project concept, through site investigation and modelling, to construction, monitoring and back analysis, allowing the whole process to be concurrently checked providing a transparent and traceable audit trail of results and associated decisions made. & 2010 Elsevier Ltd. All rights reserved.

Keywords: Rock mechanics Site investigation Modelling Design Technical auditing

1. Introduction 1.1. Purpose and principles of technical auditing It is of benefit to be able to formally audit the content of the rock mechanics modelling and rock engineering design of a project in order to ensure that all the necessary factors are included and that the technical work is correct within the limits of the site information and current scientific knowledge. The term ‘technical auditing’ is used to describe this process and the overall purposes of a technical audit are as follows: (a) to evaluate the logic of the work based on the stated objective; (b) to establish whether all the necessary physical mechanisms, variables, and parameters have been included in the relevant analyses; (c) to show that the supporting analyses are appropriate within the limits of site and scientific knowledge; (d) to consider whether the modelling and design analysis conclusions are justified in terms of the project objectives; and (e) to provide an audit information, analysis and decision trail. The key principles of an audit in general [1] are that it is made according to evidence, known criteria and the current scientific framework. Auditing involves verification by evidence and the result is an opinion based on persuasive evidence. The audit should have an independent status, be free from investigatory and reporting constraints, produce a benefit, and result in a report.

n

Corresponding author. Tel.: + 44 170 732 2819 E-mail address: [email protected] (J.A. Hudson).

1365-1609/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijrmms.2010.05.001

The audit result will always be an opinion; hence, the auditing must carry authority. These general principles also apply to the specific case of technical auditing for rock mechanics modelling and rock engineering design and have therefore been adopted here. 1.2. Information supporting the technical audit One of the key aspects of the audit is that it will depend on the information potentially available. In a previous paper [2], we noted that the overall procedure for obtaining the necessary information follows the steps described below. These begin with the choice of modelling methods, because different modelling methods require different informational inputs. 1. Choose the modelling methods according to the objectives and sub-objectives of the project, the identified features and constraints of the site, rock mass and project and applicability ranges of the modelling methods. 2. Identify the information required for the chosen modelling methods. 3. Choose the methods for obtaining the data required for the selected modelling methods from the site investigation techniques, e.g. aerial-photography, geological surveys, laboratory testing, field testing, in situ monitoring, etc. 4. Interpolate/extrapolate any missing information by using engineering experience, back analysis, data mining for knowledge and/or perform sensitivity analyses on the parameters to provide probabilistic data.

878

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

The two demonstration auditing examples that follow are: (1) the procedure for auditing a site investigation measurement— in situ rock stress (soft audit) and (2) auditing the modelling for the design of hydropower caverns at the Laxiwa site on the Yellow River in China (semi-hard audit). Whether a soft or hard audit is appropriate in any given situation will depend on the requirement and according to the descriptions in Fig. 2; both forms of audit are demonstrated in the following sections. The limited audit illustrated for the second demonstration audit for the hydropower caverns in China is termed ‘semi-hard’ because it is not possible to include the full detail given the limitations on the paper length.

5. Undertake rock engineering modelling and design and record all steps in the process.

With reference to Item 4 in the above list, an additional benefit of the technical audit is that it enables highlighting of any ‘guess work’ based on engineering experience that is potentially subject to both amendment with further experience and updating as science advances. This overall procedure was also presented within the context of two flowcharts: one illustrating the eight basic modelling methods (Fig. 1, noting that there are four methods A–D with increase in complexity from left to right, each with two submethods depending on whether the method involves 1:1 mapping or non-1:1 mapping of the project geometry); the other illustrating the seven basic steps in the rock engineering design process [2]. All the aspects of Fig. 1 can be subject to technical auditing, i.e. obtaining the site and rock mass information from site investigation, use of the eight types of modelling methods (whether used singly, in combination or in sequence), development of the initial design, construction monitoring and back analysis, leading to the final design. Similarly, the auditing is applied to the seven design steps discussed in [2], these being the project purpose, key features of the site/rock mass/project, design approach strategy, choice of modelling method and appropriate code, initial design, integrated modelling and feedback information and final design and verification/validation.

2. Demonstration example 1: the procedure for technically auditing a site investigation measurement—in situ rock stress (soft audit) 2.1. Background The validity of the modelling and design of a rock engineering project will depend on the accuracy of the supporting information

1.3. ‘Soft’ and ‘hard’ technical audits and the audit evaluation Note that the technical auditing can be ‘soft’ or ‘hard’, with the characteristics as indicated in Fig. 2. Thus, for modelling, i.e. using one or more of the eight main types of modelling methods shown in Fig. 1, either auditing type can be utilized, depending on the purpose of the auditing. The soft audit can be used initially to support the development of the modelling programme. However, the hard audit is necessary for the total audit evaluation and ability to state whether the modelling is adequate for the purpose.

'Soft' Audit

'Hard Audit'

Audit Evaluation

Obtains the basic information for establishing the essence of the problem

Obtains the detailed information on all the procedures being used

Establishes whether the modelling is adequate to meet the objectives

Ability to present what is being done

Ability to state the details of what is being done

Ability to state whether the modelling is adequate for the purpose

Fig. 2. The ‘soft’ and ‘hard’ audits and the audit evaluation.

Objective

Method A

Method B

Method C

Method D

Use of pre-existing standard methods

Analytical methods, stress-based

Basic numerical methods, FEM, BEM, DEM, hybrid

Extended numerical methods, fully-coupled models

Level 1 1:1 mapping

Precedent type analyses and modifications

Rock mass classification, RMR, Q, GSI

Database expert systems, & other systems approaches

Integrated systems approaches, internet-based

Level 2 Not 1:1 mapping

Site Investigation

Design based on forward analysis

Design based on back analysis

Construction

Fig. 1. Flowchart of rock mechanics modelling types used to support rock engineering design (from [2]).

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

concerning the rock mass properties—which comes from site investigation and back analysis. Thus, technical auditing procedures checking the correct implementation of method descriptions must be developed for the different elements of the site investigation. The demonstration example presented here is the soft audit

879

procedure for in situ rock stress estimation, chosen because rock stress is a particularly elusive and difficult parameter, yet one that is crucial as the boundary conditions for modelling. Moreover, the technical audit for this parameter indicates the style of soft auditing for all site investigation parameters.

Table 1 List of technical audit subjects to be addressed in the development of a stress estimation/measurement programme. Audit Subject Area 1: Stress measurement objective and background 1. Statement of the measurement objective: What is the purpose of the measurements? What is the accuracy expected? What confirmatory procedures are to be adopted? 2. Statement of the stress measurement background: Have the problems with in situ rock stress measurements been identified? Has a list of the problems been made? Have the most relevant literature references been identified and studied? Has the project been discussed with someone who has practical experience of measuring stresses, and with the specific method to be used? Audit Subject Area 2: stress measurement method 3. Specification of the stress measurement method: What stress measurement method is to be used? What are the physical processes involved? What influence might site conditions have on the results from the method to be used? What problems have been identified in the past? 4. Confirmation of method adequacy: Given the statements produced so far, is the stress measurement capable of measuring the required rock stress? 5. Availability of a QA procedure: Is a QA procedure available for the stress measurement method? If so, has the QA procedure been checked—for both theoretical and practical experience aspects—to ensure that it is adequate, given the objective and the known problems with stress measurements? Is the existing QA procedure adequate? If a suitable QA procedure is not available, can an adequate one be generated? 6. Stress measurement protocol: Is a protocol being developed for the use of the stress measurement method that incorporates the TA and QA aspects? Audit Subject Area 3: contractual aspects 7. Schedule and required resources: What time is available for the stress measurement works? Requirements on the field crew? Need for on-site auditing? 8. Auditing requirements and strategy: Conclude auditing requirements based on Subject Areas 1 and 2. Establish auditing strategy. Establish auditing resources 9. Roles for Client and Contractor: Responsibilities on site. Resources provided by the Client. Review and evaluation tasks Audit Subject Area 4: establishing QA procedures 10. Adaptation to the quality system of the organisation. Level of detail? Compatibility/coincidence with any overall QA system for the organisation? 11. Manufacturing or assembling of equipment. Are the parts used suitable for use? Are stress magnitudes, water pressure and water quality issues considered? Are the parts used of sufficient quality for their purpose? Will spare parts be available? Are the critical activities that may influence the quality of test results understood and sufficient procedures and quality control established? 12. Routines for storage and maintenance of equipment for stress measurements. Are maintenance procedures of the critical equipment components established? Is equipment stored in a safe way when not in use? 13. Quality control of data acquisition systems. Are calibration routines established? Are the lifetimes of the components understood? Is the software validated? 14. Establishment and maintenance of QA procedures. Is there a system to follow up on the routines and procedures applied? Are there established and maintained procedures to identify training needs, as well as to provide the training, of personnel carrying out and evaluating the stress measurements? Audit Subject Area 5: quality aspects for establishing the viability of stress measurements at a given location/depth 15. Decision on the test location/depth. Is the most recent geological information being used for judgment of the suitability of a test location/depth? Is the proposed test location/depth representative for the site/the planned project? For overcoring: are there specifications on required rock quality in the actual formation stated in advance of the measurements? For hydraulic fracturing: is the influence of any anisotropy on test results understood, and could the least anisotropic sections be chosen? For HTPF: are there suitable closed fractures available? 16. Functional testing of installation tools, etc. Are procedures followed and checklists used? 17. Procedures to install equipment at the suitable location/depth. For overcoring: what procedures are in place for drilling the pilot hole and accepting the test level? For hydraulic fracturing: what procedures are in place for controlling that the packers are placed at the chosen test level, and are relevant procedures followed and documented? Audit Subject Area 6: measurement procedures 18. Down-hole installations and measurements. What procedures are in place to ensure that the down-hole operations are fully traceable? What procedures are in place to check the actual geological conditions at the test level (e.g. inspection of overcored sample, checking impression packer result)? 19. Data acquisition What procedures are in place to check or calibrate gauges used? What procedures are in place to check hardware and software? What procedures are in place for data storage and backup? Audit Subject Area 7: stress data reduction and interpretation 20. Data recording, reliability and reduction. What procedures are in place to ensure that the data will be recorded accurately and safely? Have all the hazards with stress measurements (see Audit Subject Areas 1 and 2) been addressed? What procedures are in place to ensure that the raw data obtained are reliable? How will the data be reduced? What procedures are in place to ensure that mistakes will not occur during data reduction? Is there a protocol with a case example available for this? 21. Data interpretation How are the data to be interpreted and the trends identified? Audit Subject Area 8: continuous evaluation process 22. Procedures for on-site evaluation and draft reporting What procedures are in place for a gradually updating understanding of the results during the measurement process? What procedures are in place for a decision on continuation or termination of field works (see Subject Area 2)? Audit Subject Area 9: validation and presentation 23. Data validation Are results compatible with existing relevant data and trends at the site? Are the site conditions within the assumptions for the method used? Are the determined elastic properties of the rock realistic? 24. Presentation of stress measurement results How are the stress measurements to be presented in a clear form? Discussion of the process for uncertainty evaluation. How is the uncertainty to be presented? Audit Subject Area 10: technical auditing conclusions 25. Stress measurement adequacy Have the stress measurements been conducted adequately, given the objective (Audit Subject Area 1) and the existing scientific, practical and site knowledge? Is the documentation of the quality control during measurement, data reduction and data interpretation reliable (Subject Areas 5–9)? 26. Overall Technical Auditing statement What are the overall TA conclusions given the individual conclusions in Items 1–25 above? What recommendations are to be made concerning the work?

880

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

It is noted in the first of the ISRM Suggested Methods on stress estimation [3] that the objective of the stress estimation programme must be established, plus the ramifications of the objective. What information is required? Principal stress directions? The magnitude of one or more principal stress components? The complete stress tensor? The variation of the stress state across the site? Are general estimates required, or determination via actual measurements? Are the values required with an interpretation of the site context? What accuracy is required? How are uncertainty and spatial variability to be assessed? Is a confirmatory procedure required? Is a multiple complementary approach required with a final quantitative harmonization? Do the results need to be supported by subsequent numerical modelling? How are the results to be presented? Is strict quality control required, or is an informal approach satisfactory? All of these questions should be answered, both for the site investigation and via the audit. Thus, the auditing itself, if carried out concurrently with the site investigation can assist in planning the work. The second and third ISRM Suggested Methods concern overcoring and hydraulic fracturing and in the fourth of the ISRM Suggested Methods for stress estimation [4], a procedure is outlined for auditing in situ stress measurement. A condensed version of this procedure is presented here as the first example of technical auditing.

Province, China. There is a large cavern group for the underground powerhouse excavated in granite in mountainous terrain (Fig. 3). The project consists of the main powerhouse, auxiliary powerhouse, main transformation cavern, surge chamber, tailwater lock chamber and tailwater tunnel. The main powerhouse is 312  30  74 m3 in length, width and height with the azimuth of the axis being NE251 (Fig. 4). Detailed information on the geological conditions related to this project can be found in [5]. A 3D numerical analysis has been made to assess stability of the cavern group during excavation, optimize excavation procedure of each cavern and design of the support system for the caverns (Fig. 5). 3.2. Auditing the modelling for the Laxiwa Project design The technical auditing of the modelling for the Laxiwa Project is termed ‘semi-hard’ here because, given the constraints on paper length, it is not possible to include all the relevant detail here for the full hard audit. The main design issue was establishing a suitable cavern excavation sequence and associated support in the highly stressed granitic rock mass. The technical audit was

2.2. Auditing rock stress measurements The preceding Table 1 is condensed from [4] and has ten subject areas with twenty-six audit questions. This is a ‘soft audit’ because the audit questions are orientated towards ensuring that the overall conditions for a successful stress estimation campaign are achieved, rather than the complete detail required for a hard audit. Thus, Table 1 can be used for the planning of a stress measurement campaign, for contemporaneous auditing during the stress measurements, or for subsequently auditing a completed stress estimation programme. It acts both as a checklist guide to the key aspects and as a structure for the auditing process itself.

Fig. 3. Indication of the topography at the Laxiwa hydropower project site on the Yellow River in China.

2.3. Discussion and conclusions relating to demonstration example 1 Naturally, the audit subject areas and procedural questions in Table 1 are specific to rock stress estimation and represent a soft audit, but similar questions should be asked for all the site investigation measurements. It is easy to make mistakes when obtaining site data—for a variety of reasons: e.g., incomplete understanding of the subject matter, numerical errors, malfunctioning equipment, lack of communication on site, etc. Hence, being systematically alerted to the potential pitfalls when making site investigation measurements, as illustrated for the case of in situ stress estimation, through addressing the types of questions listed in Table 1, will prove useful to all parties involved: the site investigation contractor, the modeller, the designer and the auditor. Such auditing, or its equivalent, should be implemented for all site investigation measurements relating to a project in order to support the validity of the rock mechanics modelling and rock engineering design.

river

Yellow

Z

Y

Cavern group

X

3. Demonstration example 2: technical audit of modelling for the design of hydropower caverns at the Laxiwa Yellow River site in China (semi-hard audit) 3.1. The Laxiwa hydropower project on the Yellow River The Laxiwa hydropower station is located on the Yellow River at the boundary of Guide County and Guinan County in Qinghai

Fig. 4. Location and topography at the cavern group site, Laxiwa hydropower project.

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

881

Fig. 5. (a) 3D model for numerical analysis and (b) calculation model for excavation of the Laxiwa cavern group.

Table 2 The 11 subject headings for the semi-hard technical audit of the rock mechanics modelling supporting the cavern excavation design for the Laxiwa project on the Yellow River, China. 1. THE MODELLING OBJECTIVE The purpose of the modelling? 2. CONCEPTUALIZATION OF THE PROCESSES BEING MODELLED The sub-system(s) being isolated for study. The physical processes involved. 3. SPECIFICATION OF THE MODELLING CONTENT What are the physical variables, connecting relations, parameters, boundary conditions, initial conditions, etc. 4. MODELLING SOLUTION REQUIREMENTS What type of model output is required, given the stated modelling purpose? 5. MODELLING SOLUTION TECHNIQUE How is the required model output to be obtained? 6. NUMERICAL CODE UTILIZED Which numerical code is to be used? How do we know that the code is operating correctly? 7. SUPPORTING MODEL DATA & DATA INPUT METHOD What are the necessary supporting data? How are they to be obtained? How are they to be input? 8. MODEL SENSITIVITY ANALYSIS How does the model output depend on the model input in terms of whether a sensitivity analysis is required? 9. PRESENTATION OF MODELLING RESULTS Is it possible to demonstrate that the numerical code is operating correctly? Are the modelling results clearly presented? 10. SOURCES OF ERRORS What are the main sources of errors? 11. MODELLING ADEQUACY Does the modelling seem adequate for the purpose? Are there any problem areas? Is any corrective action required?

conducted within the eleven subject headings listed in Table 2. The individual auditing questions and the answers for the design approach used in this case are listed in Table 3 with the eleven subject headings listed under four subject areas. The audit subject areas in Table 2 are different from those in Table 1 because now we are dealing with a design problem, rather than site investigation measurements. However, the same logic is followed in Table 3 beginning with the objective, Subject Area 1, and following the key steps through to the conclusions, in this case the modelling adequacy, in Subject Area 4. Subject Area 2 in Table 3 concerns the conceptualization of the processes being modelled, the specification of the modelling content, the modelling solution requirements and the modelling solution technique. This is a crucial area because, as indicated by the associated sub-questions in Table 3, there needs to be justification of the exact modelling approach in terms of comprehensiveness of the physical variables represented and the specification of all the conditions.

Subject Area 3 in Table 3 concerns the modelling technique itself: the numerical code used, the supporting data, sensitivity analyses and presentation of the results. Because our rock mechanics and rock engineering knowledge is continuously increasing and numerical models are continuously being improved, the auditing questions themselves may need to be adjusted to capture the critical aspects in this subject area. The key issues are ensuring that there is a check on the appropriateness of the model, that there have been checks on how the model is used, ensuring that uncertainties are captured by sensitivity studies as far as is reasonable, and that the modelling results appear valid. Subject Area 4 in Table 3 covers model adequacy more explicitly through the identification of actual or potential errors and the possible need for corrective action. In the current absence of any internationally agreed check on the rock engineering design process (as occurs in other engineering areas such as aircraft design), the auditing procedure illustrated in this demonstration example 2 shows how the answers to the questions listed in Table 3 enable confirmation that an adequate procedure is being followed or has been followed. The auditing can be used during or after the modelling activity. Needless to say, contemporaneous auditing is preferred so that any corrective action can be immediately implemented. 3.3. Comments on the Laxiwa case example It was noted earlier that the technical auditing of this case example has been termed ‘semi-hard’ because it is has not been possible to include all the relevant details in the paper. However, the audit, via the eleven subject areas and thirty-eight questions with their detailed answers in Table 3, illustrates the style of the hard audit and how the answers reveal the procedures used and their suitability in determining the cavern excavation sequence and appropriate support. To demonstrate this example as a truly hard audit would require illustration of the more penetrating investigation of the correctness of the supporting data, numerical analyses, etc. Noting that, in this case, the manner in which the answers to the auditing questions are given has been left open, to make the audit more efficient the form of the answers should be specified in more detail, e.g. whether a narrative or numerical answer is required to a particular question.

4. Conclusions In order to ensure that the procedures used in rock mechanics modelling and rock engineering design are appropriate for the

882

Table 3 Answers to the semi-hard auditing questions for the semi-hard auditing of the modelling for the design of the cavern excavation sequence at the Laxiwa hydropower site in China. Auditing component

Associated questions

Subject Area 1: modelling objective—establishing the purpose of the work 1. THE MODELLING OBJECTIVE 1.1 Has the modelling objective been The purpose of the modelling? clearly established?

Answers

The modelling objective for comprehensive analysis of the stability of the large cavern group in granite under high in situ stresses has been established. It focusses on:

1.2 How will it be known when the modelling work is completed?

Subject Area 2: modelling concept—describing the modelling concept and content 2.1 What rock mass systems are being 2. CONCEPTUALIZATION OF THE considered? PROCESSES BEING MODELLED The sub-system(s) being isolated for study. The physical processes involved. 2.2 What are the main physical processes being modelled?

2.3 What is the changing independent variable?

3. SPECIFICATION OF THE MODELLING CONTENT What are the physical variables, connecting relations, parameters, boundary conditions, initial conditions, etc.

The stability of the large cavern group in granite under high in situ stresses was analyzed and verified step-by-step from top to bottom. At the design stage of the project, the stability of the cavern group was analyzed for the designed excavation procedure and support parameters using the obtained information and verified by observation after excavation of the first level. The observation and measured deformation and damage zone were used to recognize the parameters of the model. The recognized parameters were used to analyze the stability of the cavern group induced by the remaining excavation steps (i.e., excavation at the lower levels). The stability estimation for the large cavern group after excavation at all levels (steps) using the recognized parameters should be in good agreement with the observations. The rock mass system is intact, massive granite with some faults and joints

The physical process of damage evolution of the surrounding rocks induced by excavation subject to high in situ stresses. The character of the deformation and failure in the brittle rock during excavation and ductility at high confining pressure in the long term are considered during the analysis. The effects of fractures, faults and joints, are considered. The effectiveness of the support is also simulated. Essentially, the removal of the rock during cavern excavation, i.e. incremental m3 of excavation.

2.4 How is the system perturbed so that the mechanisms are initiated?

As in 2–3 above. The damage zone of the surrounding rocks induced by the current excavation step would be affected by further excavation of the cavern group at lower levels. Higher and higher sidewalls of the caverns will be formed gradually step-by-step from top to bottom. The stability of a cavern or tunnel would also be affected by excavation of adjacent caverns or tunnels.

3.1 Listing of the physical variables

Displacement, stress, local energy release rate, plastic zone, Young’s modulus, Poisson’s ratio, tensile strength, shear strength, compressive strength, peak cohesive strength, residual cohesion, plastic strain for cohesive strength degradation, plastic strain required for the full frictional strength mobilization, friction angle

3.2 Listing of the THM couplings?

The analysis only includes the M component: the deformation and fracturing process of granite subject to high in situ stresses. There is no significant thermal or water flow processes at the site.

3.3 Is the model 1-D, 2-D, 3-D or some combination?

3D simulation for the whole rock mass body, combined with 2D simulations for local key sections

3.4 Are you modelling a continuum or a discontinuum?

The rock mass is treated as a continuum

3.5 Specification of the boundary conditions

1. The axes system is defined. 2. The calculation area includes the boundary of 210 m from left, right, downstream and upper stream of the boundary of the main powerhouse, respectively, 280 m from bottom of main powerhouse and top surface of the earth, respectively.

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

1. estimating the stability of the cavern group, including deformation values and their distribution, plastic zone depth, stress concentration zones, etc.; 2. assessing failure risks induced by excavation of the cavern group at lower levels; 3. optimizing the excavation procedure of the cavern group; 4. establishing the necessary strengthening reinforcement locally for the large cavern group induced by excavation. The monitored displacement increase and excavation damage zone induced by excavation at the upper levels can be used to recognize model parameters for input to the analysis.

3. Excavated cavern group given. 4. Excavation restriction conditions: step-by-step from top to bottom. Excavation procedure was optimized to obtain minimal damage of the surrounding rock induced by excavation. 5. Support installation: will be finished when the surrounding rock has sufficient deformation, but has not failed, and it is convenient to implement.

5. MODELLING SOLUTION TECHNIQUE How is the required model output to be obtained?

1. In situ stresses: the measured values from borehole measurements with estimated orientations. 2. Geography and topography obtained from geological survey. Geological conditions: F7, F3, HF2, HF8, HL2, L28, f11 (Chinese system), are included in the calculation model. 3. The cavern-peripheral rock is all considered as category II granite. 4. The laboratory tests and engineering analogies indicated a (strain-dependent cohesion weakening)-(friction strengthening) model can be used as the constitutive model for granite subjected to high in situ stresses. 5. Three tentative excavation bench heights, 8, 11, 15 m 6. Three tentative support schemes 1, 2 and 3. 7. The measured excavation damage zone at an exploration tunnel which is adjacent to main cavern. 8. Mechanical parameters for rock masses and joints obtained at the design stage.

3.7 How is the final condition established?

1. The parameters of the strain-dependent cohesion weakening-friction strengthening model were recognized with input from the measured excavation damage zone and deformation induced by excavation at several upper levels using the genetic algorithm – support vector machines – FLAC algorithm. 2. The strata conditions with the categories for the surrounding rocks were input for stability analysis of the cavern group induced by excavation at lower levels. 3. Height of the excavation bench for the cavern group is firstly optimized using the obtained model parameters and FLAC analysis 4. Excavation procedure with nine steps for the main cavern was optimized together with the support scheme using the proposed PSO-FLAC algorithm. 5. Stability analysis of the cavern group after finishing excavation was conducted with input from the actual excavation procedure and the resulting rock response.

4.1 What is the required model output?

It includes: 1. Displacement, stress, strain, local energy release rate, plastic zone of the surrounding rocks, from numerical analysis; 2. The optimal excavation procedure and support parameters, distribution of the deformation field, stress and plastic zone distributions, and local energy release rate of the surrounding rocks; 3. Appraisal of the overall and local stability of the cavern group and estimation of potential failure risks, depth and locations in the surrounding rocks and their causes; and 4. Suggestions for local enhancing reinforcement and an effectiveness appraisal.

4.2 Does the model output match the modelling objectives?

Yes

5.1 In principle, how is the model output to be obtained: one code, one set of data, one run?—or a suite of numerical experiments?

The rock mass was in equilibrium in a state of three dimensional stress and this was changed at the cavern surfaces by excavation to a state of two dimensional stresses (stress perpendicular to the cavern wall¼ 0). The excavation results in concentrations of stress, release of some stored energy and potentially fracturing initiation, propagation to the final state of the surrounding rocks. At each excavation step, the calculation is carried out iteratively and reaches a new balance. And then the calculation for the next excavation step is performed.

5.2 Are any quality control checks in place? Checking the input data have been entered correctly, validation against known solutions, independent duplication of runs?

Yes. The iterative calculation is convergent. The output results, such as displacement, stress, local energy release rate, plastic zone, are understood from the mechanisms involved, experience of similar projects and verified by measurement afterwards. Input of parameters is checked before the calculation by back analysis on the previous monitored results. The software is verified by using known case studies examples.

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

4. MODELLING SOLUTION REQUIREMENTS What type of model output is required, given the stated modelling purpose?

3.6 Specification of the initial conditions

Grey correlation analysis was used to check the applicability of the recognized parameters. The calculated depths of the excavation damage zone at another four boreholes at different sections, whose measurements were not used for parameter recognition, agreed well with the measurements. The calculated displacement-time series after excavation of the first seven steps of the main underground powerhouse using the recognized mechanical parameters are in good agreement with the measurements, both for the monitoring point A4-7 (whose monitored displacement was used for back analysis) and A5-4 (whose monitored displacement was not used for back analysis). 883

884

Table 3 (continued ) Associated questions

Answers

6. NUMERICAL CODE UTILIZED Which numerical code is to be used? How do we know that the code is operating correctly?

6.1 Which numerical code is to be used?

A software code using the FISH programming language has a function to calculate the local energy release rate for a straindependent cohesion weakening-friction strengthening model, genetic algorithm search, and back analysis. The programme is combined with FLAC3D

6.2 Why is that code being used?

FLAC3D is a popular numerical analysis programme in rock engineering but a new feature was developed, i.e., a strain-dependent cohesion weakening-friction strengthening model, genetic algorithm search, and back analysis. The new functions are coded using the FISH programming language and combined with FLAC3D FLAC3D originates from ITASCA, Minneapolis, USA. The new functions were coded using the suggested formula by the code developers.

6.3 Where did the code originate from?

Subject Area 3: modelling technique 7. SUPPORTING MODEL DATA AND DATA INPUT METHOD What are the necessary supporting data? How are they to be obtained? How are they to be input?

8. MODEL SENSITIVITY ANALYSIS How does the model output depend on the model input in terms of whether a sensitivity analysis is required?

6.4 How has the code been validated?

Yes. The new functions, such as the new appraisal index and new constitutive model, were validated by using two existing case studies: the AECL mine-by tunnel in Canada; and the Taipingyi hydraulic tunnel in China. The back analysis function was validated by performing back analysis of mechanical parameters for the permanent shiplock slope at the Three Gorges Project, and Shuibuya hydraulic underground powerhouse, China.

7.1 Listing of type and justification of boundary conditions

1. 2. 3. 4. 5.

7.2 Listing of input data with source of the data and justification.

1. Layout of project and location from the designer. 2. Distribution of strata and faults/joints, contour lines for topography, from geological survey report. 3. In situ stresses: the underground powerhouse cavern group is located in a high in situ stress field. The three dimensional geostress measurements indicated that the maximal principal stress was  22 to  29 MPa (compression negative) and dipping towards the Yellow River gorge, the intermediate principal stress is about  15 MPa and dipping towards the mountain. The minimal principal stress is almost vertical with a value of about  10 MPa. The geo-stress field from the geo-stress measurement report and the directions of the three principal stresses were plotted on a stereographic projection to check whether they are mutually orthogonal. 4. The characteristics of the cavern surrounding rocks were revealed during construction. 5. The monitored displacement and excavation damage zone of an exploration tunnel were obtained from the designer for recognition of model parameters for the design stage. 6. The monitored displacement and excavation damage zone of the main cavern at the upper levels was obtained from the designer (Table 9) for recognition of model parameters for the construction stage. 7. The optimal excavation procedure 9b and support scheme 1 were used for the stability analysis of the cavern group for the design stage. 8. The actual excavation procedure 9b and support scheme 1, with the recognized model parameters for the stability analysis of cavern group after finishing all excavation steps.

7.3 Do the data have to be adjusted before being input?

Yes. The surrounding rocks at the downstream sidewall of the main cavern were considered as category II, but were revealed to be category III after the first step excavation. The corresponding mechanical parameters such as Young’s modulus, tensile strength and shear strength had to be changed according to different damage zones from wall to inside of the surrounding rocks. The calculated results were different for the downstream sidewall of the main cavern. Another small data adjustment relates to the mechanical parameters for the strain-dependent cohesion weakening-friction strengthening model. During the design stage of the project, the parameters were established by using the monitored excavation damage zone at an exploration tunnel which is adjacent to the main cavern. However, the parameters of the model were established again using the monitored displacement and excavation damage zone induced by excavation of the main cavern at upper levels and used to analyze stability of the cavern group for excavation at the lower levels. The deformation of the surrounding rocks depends on their mechanical parameters. For example, deformation at the downstream sidewall of the main cavern was generally smaller than that of the upper stream sidewall of the main cavern when the surrounding rocks are considered as category II from data in the design stage of the project. When data from the construction stage was used, the surrounding rock at the downstream sidewall of main cavern was changed to category III and the results were reversed. Yes. A sensitivity analysis was conducted to choose which parameters should be back analyzed. Errors of  30%,  15%, 0, 15%, and 30%, respectively, were given to the parameters to be analyzed.

8.1 How does the model output depend on the input parameter values?

8.2 Is a sensitivity analysis being conducted? If so, what type of analysis? Processes, mechanisms, parameters, boundary conditions, couplings, etc.

In situ stress, back analysis based on measured values at boreholes. Deformation and damage zone measured at exploration tunnel and cavern at upper levels. Geometry, faults, joints, and strata layers measured from geological survey Layout of cavern group from the designer. Observation of failure and deformation phenomena of surrounding rocks from the construction monitoring.

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

Auditing component

9. PRESENTATION OF MODELLING RESULTS Is it possible to demonstrate that the numerical code is operating correctly? Are the modelling results clearly presented?

The results indicated that parameters such as Young’s modulus, peak cohesion, plastic strain for cohesive strength destruction, plastic strain required for the full frictional strength mobilization, friction angle of the rock mass are sensitive to the monitored displacement and excavation damage zone and could be recognized by using values of the monitored variables.

9.1 Is it possible to demonstrate that the numerical code is operating correctly?

Yes. It can be demonstrated in several ways. One is comparison of the predictions using the established parameters for the model with measurements not used to recognize the parameters. Another is the analysis results at locations having stress concentration, large stress relaxation, high local energy release rate, large tensile strain, potential failure, etc., for the excavation of the cavern group at lower levels were mostly verified by observation afterwards.

9.2 Is it possible to show that the supporting data are reasonable assumptions for a rock mass?

Yes. The predictions from the strain-dependent cohesion weakening-friction strengthening model and Mohr-Coulomb were compared with the measured excavation damage zone indicating that the model used was correct.

9.4 Does the presentation of the modelling results link with the modelling objective?

The correctness of the updated data of Young’s modulus, tensile and shear strength for the downstream sidewall of the surrounding rocks was verified by comparing their predictions with the measurement afterwards. The distribution of displacement, local energy release rate, plastic zone, and stresses were drawn respectively for the central cross-sections of every generator unit, horizontal and axial vertical sections of the cavern group. The differences in displacement, local energy release rate, plastic zone, and stresses at key locations were compared in graphical mode and tables were constructed for different excavation procedures, support schemes, with and without locally strengthening support, and with and without ‘bulgy’ bench. Yes. For examples, tensile strain plots show where there would be tensile failure and slabbing failure. The local energy release rate figures indicate the location and depth of brittle failure of the surrounding rocks. The plastic zone figures show where there are tensile and shear failure elements with different colours in the rock mass.

10.1 Have you already corrected any errors?

Yes. The category of the surrounding rock on the downstream sidewall of the main cavern was corrected from category II to category III.

10.2 List the sources of potentially significant errors.

1. 2. 3. 4.

10.3 Do any of the potentially significant errors invalidate the modelling objective, concept and conclusions?

No, but changes were necessitated in

11.1 Do all the previous questions indicate that in principle the model is adequate for the purpose.

Yes

11.2 If not, list the problem areas.

N/A

11.3 What corrective action is required?

None

11.4 Does this semi-hard audit have to be repeated after corrective action has been taken?

No

9.3 How are the modelling results to be presented?

Subject Area 4: Modelling Adequacy 10. SOURCES OF ERRORS What are the main sources of errors?

11. MODELLING ADEQUACY Does the modelling seem adequate for the purpose? Are there any problem areas? Is any corrective action required?

Geological conditions, strata, joints and fractures, etc., to be simulated; Zoning of damaged surrounding rocks induced by excavation; Difference between simulated with actual excavation procedure; Difference between simulated and actual support time.

(1) Category of surrounding rocks. (2) Support time and effectiveness.

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

8.3 How are the results of the sensitivity analysis to be summarized?

885

886

J.A. Hudson, X.-T. Feng / International Journal of Rock Mechanics & Mining Sciences 47 (2010) 877–886

purpose, the technique of technical auditing can be used. There is a requirement for a technical audit if there is a need to establish the reliability and credibility of information, or if there is a public interest dimension requiring public accountability. Greater client and public confidence is achieved if the supporting studies are presented in a transparent and traceable manner with a full audit trail of work done and decisions made. The technical audit can be conducted in a ‘soft’ or ‘hard’ fashion. The soft audit obtains the overall information and concludes with a presentation of what is being done. The hard audit obtains the more detailed information on all the procedures being used, sufficient to establish whether these procedures are adequate to meet the objective. All aspects of the modelling and design can be technically audited, from establishing the objective, the site investigation, modelling methods, initial design, construction monitoring, back analysis and final design. The technical auditing procedure can be applied ‘before and during the event’ to plan and guide the work; it can also be used ‘after the event’ to audit a modelling or design exercise that has already been completed. In both cases, the work is rendered more transparent. The ‘before and during the event’ auditing is preferred because this enables the identification of problems in the on-going work and hence suitable changes to be made, resulting in a greater chance of the modelling and design work being correct. The first technical auditing demonstration was a soft audit example—the procedure for technically auditing the estimation of the in situ stress state during a site investigation. Although the audit questions presented in this demonstration are specific to rock stress estimation, the audit could easily be adapted to other site investigation measurements. Being systematically alerted to the potential pitfalls when making site investigation measurements, through addressing the types of questions listed in Table 1, will prove useful to all parties involved: the site investigation contractor, the modeler, and the designer. The second technical auditing demonstration was a semi-hard audit—the procedure for technically auditing the design of the excavation sequence for a series of hydropower caverns in highly stressed granite at the Laxiwa site on the Yellow River in China. The eleven subject areas and thirty-eight questions with their detailed answers illustrates the style of the audit and how the answers reveal the procedures used and their suitability in determining the cavern excavation sequence and appropriate support. An auditing conclusion was that, to make the audit more

efficient, the form of the answers should be specified in more detail, e.g. whether a narrative or numerical answer is required to a particular question. From the technical auditing descriptive text and the two demonstration examples presented in this paper, the benefits of auditing rock mechanics data, modelling and rock engineering design are apparent: interacting with the technical audit provides guidance, a check on procedures and supporting data, identification of problems and an independent assessment of the work. It will have been noted that the two demonstration auditing examples in the paper contain different subject headings and different detailed questions—because they have been tailored to the type of work being audited. Future work will be directed towards the development of technical auditing frameworks that can be used for the full range of rock mechanics modelling and design activities.

Acknowledgements Financial support from the National Natural Science Foundation of China under grant no. 50539090 was received for part of the study described in this paper. Yao Shuanxi, Ren Zongshe, Yang Cunlong, Shi Guangbin, Song Yongjie and Liu Jian provided help with the calculations for the project. They are all gratefully acknowledged. The authors also appreciate anonymous reviewing comments, which have significantly helped to improve the quality of the presentation by alerting us to the need for a variety of clarifying explanations—which have been included in this final version of the paper.

References [1] Giove FC. The essentials of auditing. Research & Education Association; 1998. [2] Feng XT, Hudson JA. Specifying the required information for rock engineering modelling and design. Int J Rock Mech Min Sci 2008;47:179–94. [3] Hudson JA, Cornet FH, Christiansson R. ISRM suggested method for rock stress estimation—Part 1: strategy for rock stress estimation. Int J Rock Mech Min Sci 2003;40:991–8. [4] Christiansson R, Hudson JA. ISRM suggested method for rock stress estimation—Part 4: quality control of rock stress estimation. Int J Rock Mech Min Sci 2003;40:1021–5. [5] Feng XT, Jiang Q. Intelligent feedback analysis on the stability of a large cavern group in brittle rock under high geo-stress conditions. Int J Rock Mech Min Sci 2009; submitted for publication.