an alternative theoretical methodology for monitoring the risks of co2 leakage from wellbores

The approach proposed in this paper, for assessing the risk of CO2 leakage, modifies Hudson’s methodology by inclusion of a concept known as Incident Potential Matrix IPM instead of the

Energy Procedia 00 (2008) 000–000

Energy Procedia www.elsevier.com/locate/XXX

GHGT-9

An Alternative Theoretical Methodology for Monitoring the Risks

of CO 2 Leakage from Wellbores

Jose Condora,b*, Koorosh Asgharib

a Energy Technology Innovation Policy, John F Kennedy School of Government, Harvard University 79 JFK St Cambridge, MA 02138, USA

b Faculty of Engineering, University of Regina 6 Research Dr Regina, SK S4S 7J7, Canada

Elsevier use only: Received date here; revised date here; accepted date here

Abstract

This paper proposes an alternative theoretical methodology to evaluate the risks of CO2 leakage from reservoirs using a stochastic approach The methodology suggested here makes use of three main concepts:

− Features, Events and Processes (FEPs)

− Interaction Matrix, IM

− Stochastic Representation

Both, FEPs and Interaction Matrix have been introduced by other researchers but for different objectives The methodology that

is proposed here modifies the original concept of Interaction Matrix in such a way that it may produce probabilistic results as outcome A practical example is given at the end of this paper

© 2008 Elsevier Ltd All rights reserved

Keywords: Geological Storage; Risk Assessment; Wellbore Stability; Leakage Scenarios; Safety Criteria; EOR

1 Introduction

Several papers have been published in recent years with the objective of evaluating the risk of CO2 leakage from reservoirs[1-5] Some of these researchers have considered the complete geo-sequestration systems, which exhibit high degrees of complexity A geo-sequestration system is a complicated group of elements that has to be modelled

in such a way that represents the real system as closely as possible

Quintesa elaborated a database with all the applicable Features, Events and Processes (FEPs) for the safe long term storage of CO2[6-8] It was originally proposed as a screening and monitoring tool for the radioactive waste management[9] This database constitutes an excellent source of information, but it does not provide a structured methodology to evaluate the risks of CO2 leakage

* Corresponding author Tel.: +1-617-496-2705; fax: +1-617-496-0606

E-mail address: Jose_Condor@ksg.harvard.edu

c

Energy Procedia 1 (2009) 2599–2605

www.elsevier.com/locate/procedia

doi:10.1016/j.egypro.2009.02.026

In 1992, John Hudson published the Rock Engineering Systems explaining in a practical way the value of the Interaction Matrix for the mines engineering[10] His work was useful later on to model engineering systems for storing radioactive waste materials His approach is simple, but practical and basically consists of what the author defined as Interaction Matrix, Expert Semi-Quantitative (ESQ) Code and the Cause-Effect Plot

The approach proposed in this paper, for assessing the risk of CO2 leakage, modifies Hudson’s methodology by inclusion of a concept known as Incident Potential Matrix (IPM) instead of the ESQ-Code The inclusion of the IPM concept in the interaction matrix may allow stochastic modelling by the use of probabilistic density functions Also, the interaction matrix can represent different scenarios that make it even more interesting for the modelling purposes

in the monitoring phase of CO2 storage projects

The first part of this paper will refer to the definitions of Interaction Matrix, Incident Potential Matrix, and the Cause-Effect Plot The second part will present an example applicable for evaluating the risk of CO2 leakage through the wellbores that are used for geological storage of CO2

2 Concept of Interaction Matrix [10]

The Interaction Matrix is a square matrix with a main diagonal containing the components of the system These components are called leading diagonal elements or LDEs The rest of the cells define the interactions and are called off-diagonal elements or ODEs By convention, the interactions between the LDEs and ODEs follow the clockwise direction as it is shown in the Figure 1 In this figure, the main diagonal (dark blue) contains the elements or components of the system and the remaining cells (light blue) define the interactions produced among them

The interactions for a geo-sequestration system are the result of physical and chemical processes represented by variety of parameters It must be noticed that the matrix is not a symmetric one, which means that Interaction of A to

B is not equal to Interaction of B to A Furthermore, certain interactions may not be considered important for the model and then their cells are represented as empty

3 Concept of Incident Potential Matrix, IPM

This matrix is mainly used in safety management in order to evaluate risks using a qualitative approach In the IPM, the risk can be defined deterministically as the product of exposure and severity

Severity Exposure

Figure 1 A 3X3 Interaction Matrix

Element A Interaction

A to B

Interaction

A to C

Interaction

B to A Element B

Interaction

B to C

Interaction

C to A

Interaction

C to B Element C

The IPM is relatively easy to use Its vertical axis corresponds to the exposure or likelihood of occurrence and its horizontal axis represents the severity The intersection between the exposure and severity produces the risk (Figure 2) The right criterion is extremely important in defining the risk A group of experts must assume the values of severity and likelihood

Consider for instance the case of an undetected geological fault that can be reactivated The probability of occurrence (exposure) may be very low (A), but the severity can be catastrophic (4) if the location is close to a populated area With these two data, the corresponding value for risk would be A4 (low risk)

The colour code for these four kinds of risks allows an easy visualization of the critic interactions In this proposal, four risk categories are identified The very low risk is assigned a value of 1; the low risk, 2; medium risk, 3; and the high risk, 4 The value of zero (0) is applicable when there is absence or undetected risk These values (Table 1) are the input data which later on will define the Cause-Effect plot

Table 1 Colour Code for Risk Evaluation

Priority

0 Nil No identified risk

1 Very Low Present Interaction – cannot be considered in the initial evaluation, but it has the

potential of affecting the system Little influence in other parts of the system

2 Low Important Interaction – part of the initial evaluation Limited or uncertain influence

through this interaction to other parts of the system

3 Medium Very Important Interaction – part of the initial evaluation They have influence in

other part of the system

4 High Critic Interaction – Part of the initial evaluation High probability of influencing other

parts of the system

Expos

High

Medium

Low

Light (1)

Serious (2)

Major (3)

Catastrophic (4)

Multi Catastrophic (5)

High Risk = 4

Low Risk = 2

Exposure or

Likelihood

(A - E)

Potential Severity (1 - 5)

Medium Risk = 3

Prevention

Very Low Risk= 1 Figure 2 The Incident Potential Matrix, IPM

4 Concept of Cause-Effect Plot [10]

Once the values for the respective process have been given in the matrix, the next step is to find some mechanism

to obtain a quantitative evaluation This is possible by the use of the Cause-Effect Plot The basic idea is to sum all the horizontal and vertical values of the cells corresponding to the interactions of an element The horizontal values are known as causes and the vertical, effects Figure 3 represent the generation of the Cause-Effects coordinates Figure 4 represents the generation of a Plot using the Cause-Effects coordinates The Plot has equidistant lines in

an angle of 45º to the axis which represents the intensity of the element, and the perpendicular lines to them represent the domain of the element By using the Cause-Effect Plot it is possible to assign numerical values of

intensity and domain of an element E i in the geo-sequestration system Both, the intensity and domain are defined by simple equations:

2 :

2 :

Figure 3 Cause-Effect Coordinates

Figure 4 Intensity and Domain of Parameters

5 Example of the Methodology

The proposed Interaction Matrix for Long Term Wellbore Stability is made up of fifteen (15) elements (LDEs), but for this example, only six (6) of the elements are considered It is important to emphasize that each geo-sequestration system has a different matrix depending on its particular features Also the processes and events may vary depending on the scenario analysis, which means that the matrix may take into account different alternatives when modelling[11] Here is where the group of experts play a crucial role in defining the interactions and the values for the incident potential matrix

Ei

Cause

2 / ) ( C − E

2 / )

Intensity

Equidistant intensity lines

Equidistant domain lines

Column : Influence of other elements

on

Rows :

Influence of

on other

elements

Elements in the

main diagonal Interactions in

the rest of the cells

Pi columns in j

∑

) (

Pj

raws in i

∑

)

(

(CAUSE)

(EFFECT)

The examples showed in the Figures 5 through 7 illustrate the three main stages in evaluating the risks of CO2 leakage from wellbores using this methodology, based on Interaction Matrix, Incident Potential Matrix and the Cause-Effect Plot

Figure 5 is a matrix containing six elements and consequently thirty potential cells for interactions Each cell may have none or several interactions and these interactions should come from the list of FEPs

Figure 6 is the same interaction matrix after evaluated by an Incident Potential Matrix The code colour constitutes an easy way to see the critical interactions The numbers in the right and bellow the matrix are resulted from the summation of the values of the interactions in the horizontal rows and vertical columns, respectively These numbers are the coordinates that provide the input for the Cause-Effect Plot

Finally, Figure 7 plots all the elements of the system identifying their domains and intensities In this particular example, it is the cement plug which holds the highest values for domain (9.90) The highest values for intensity correspond to both cement plug and water composition (16.97)

The interpretation for this particular example is that under certain assumptions, cement plug constitutes the element in the geo-sequestration system with the highest risk for CO2 leakage This information can be used for the initial evaluation of the system and consequently for devising mitigation plans to reduce their risks Also, later on, this same matrix may be used during the monitoring phase

Figure 5 Section of an interaction matrix

(The numbers in the brackets corresponds to the names of the cells and not their numerical values)

Figure 6 The Interaction Matrix after giving numerical values to the interaction cells The red numbers (right and bellow the matrix)

correspond to the sum of the horizontal and vertical values of the interaction cells These numbers are the coordinates for the

Cause-Effect Plot (Figure 7)

Figure 7 The Cause-Effect Plot for the Interaction Matrix In this particular example, is the element cement plug which has the

highest values for intensity and domain and consequently it is the element in the geo-sequestration system with the highest risk to

provide a pathway for the CO 2 leakage

6 Conclusions and Recommendations

This methodology has a great potential of being used in the probabilistic risk assessment and monitoring phases

of wellbores and other components of geo-sequestration systems used for geological storage of CO2 Its major advantage consists in its simplicity and practical results

Further studies may include the stochastic representation in the Incident Potential Matrix In such a way, instead

of having a fixed number for risk, probability density functions can incorporate the uncertainty

7 Acknowledgements

The funding for this work was provided by Natural Resources Canada, NRCan, T&I projects, and the Petroleum Technology Research Centre, PTRC and the University of Regina

8 References

[1] Nordbotten, J.M., Celia, M.A., Bachu, S., Dahle, H.K Semianalytical Solution for CO 2 Leakage through an Abandoned Well Environmental science & technology 2005 Jan 15;39(2):602-11

[2] M.A Celia, S Bachu, J.M Nordbotten, S.E Gasda, H.K Dahle Quantitative Estimation of CO 2 Leakage from Geological Storage: Analytical Models, Numerical Models, and Data Needs Paper 228, presented at 7th

International Conference on Greenhouse Gas Control Technologies

[3] F.J Moreno, R Chalaturnyk, J Jimenez Methodology for Assessing Integrity of Bounding Seals (Wells and

Keith, D.W & Gilboy, C.F.), vol 1 pp 731–740, Oxford, UK: Elsevier

[4] K Tammaoto, O Kitamura, K Itaoka, M Akai A Risk Analysis Scheme of the CO 2 Leakage from Geological Sequestration Presented at 7th International Conference on Greenhouse Gas Control Technologies

[5] K Yamamoto, K Itaoka, C Yoshigahara, M Akai Simple Estimation Methodology of Leakage from Geologic

[6] Quintessa Limited CO 2 FEP Database Available in internet [www.quintessa.org/co2fepdb]

[7] Maul, P, Savage, D, Benbow, S, Walke, R, Bruin, R Development of a FEP Database for the Geological Storage of Carbon Dioxide Presented at 7th International Conference on Greenhouse Gas Control Technologies [8] The Netherlands Institute of Applied Geoscience TNO – National Geological Survey Risk Assessment using FEPs May 2003/7.53

[9] The Swedish Nuclear Fuel and Waste Management Co., SKB [www.skb.se]

[10] John A Hudson Rock Engineering Systems: Theory and Practice Ellis Horwood, 1992 University of

Michigan

[11] Swedish Nuclear Fuel and Waste Management Co Project SAFE Scenario and system analysis Report

R-01-13