Geological risk assessments have been routinely carried out for exploration and development in oil and gas since 1990s. Because of our lack in sufficient information and knowledge of subsurface resources, the geological risk is quantitatively evaluated using probabilistic techniques. Even though the methodologies of risk assessments look well established already, some issues still remain in their actual execution and application and three key issues are highlighted here: geological chance of success, cognitive biases and post-well evaluation.

Geological chance of success may be regarded as relatively a simple concept of risk evaluation, but its proper understanding and implementation canʼt be easily achieved as it is not an absolute number and its assessment depends on quality/quantity of information and technical skills/experiences of evaluators.

As subsurface evaluation relies on fairly limited information under great uncertainties, it is very susceptible to cognitive biases including anchoring, availability heuristic, conjunction fallacy, overconfidence and hindsight bias.

Post-well evaluation is a crucial part of the risk analysis and required to be conducted so as to improve the quality of risk evaluation.

This paper was created to make a record of the lecture in the exploration technology symposium, June

2017.

In this lecture, uncertainties in geophysical evaluation were highlighted when making the geological risk assessment and estimating prospective resources. In this paper, recent technologies in the seismic industry are not described. But, some of good examples are provided for geophysicists to carefully evaluate the uncertainties attributed to the seismic data and support data such as wireline data and seismic velocity data. 1）The uncertainties in the seismic imaging caused by the statics correction and by lateral thickness changes of high velocity sediment, 2）the uncertainties in horizon picking on the laterally variable wavelets by changing fluid contents in the reservoirs, 3）the quality of the support data, 4）the uncertainties and errors in the depth conversion using the layer cake velocity model, and 5）utilizing multiple seismic attributes applicable to predict good reservoirs should be thoroughly investigated in conjunction with not only mathematical solution, but also the geoscientific solution including the regional geological setting.

Also it was highlighted concerning the adjustment of geological “Chance of Success” in accordance with “Direct Hydrocarbon Indication”.

The development of risk and uncertainty evaluations on exploration prospects were initiated by major oil companies in 1960ʼs, which are now widely used in most of oil companies over the world. Dr. Peter Rose is one of main contributors to compile and introduce the methodology widely to oil industry. There are a few points（ rules） in his methodology; 1） Parameters related to Gross Rock Volume （GRV） such as accumulated area and net pay thickness should be given as log-normal distribution, so that calculated resources also have log-normal distribution, 2） Geological Success should be given as Chance of Success （CoS） corresponding to the probability that the minimum （P99） of resource distribution occurs.

In this paper, these points （rules） in the Roseʼs method are discussed to be convinced that it is not always necessary to stick on them, since oil and gas accumulated volume are better evaluated than before. If accumulated volume can be evaluated correctly, the resource distribution is not necessary as log-normal distribution. In addition, giving only the Geological Success becomes valid, if the resource distribution can be modified to include all the uncertainties as the results of the integration with accumulated volume.

Therefore, the key to improve the Roseʼs method is to be integrated with petroleum system analysis, especially for the evaluation of oil and gas accumulated volume by geochemistry and seismic attribute analysis. In second part of this paper, oil and gas migration efficiency that is the key parameter to calculate accumulated volume by static simple method is discussed in detail. Finally, the ratio of oil and gas accumulated volume against generated volume that corresponds to the migration efficiency is demonstrated using actual cases in petroleum systems and sedimentary basins over the world.

The greatest challenge in petroleum migration risk analysis arises from the fact that specifying migration pathways is extremely difficult by using currently available data such as 3D seismic, due to their insufficient resolution. Utilizing petroleum play analogues as indirect data is, therefore, one of the important approaches to improve the analysis.

An example from UK North Sea pre-Cretaceous plays was shown to demonstrate validity of the approach. Historical exploration well data drilled in the common petroleum system or plays were collected, and the relationship between migration-related parameters and drilling results was statistically analyzed. It revealed that migration style, reservoir age and migration distance are the parameters most strongly related to successful or failed migration. Those statistical analysis results can be used as reference to determine migration chance factor for the plays, which subsequently helps to determine that for undrilled prospects.

Petroleum migration risk analysis by using play analogues must be combined with other methods such as well/seismic shows analysis and petroleum system modeling. Systematic and integrated approach must lead to better understanding and risking of plays/prospects, and new insight into further exploration potentials.

Petrophysical interpretation result is essential information for subsurface evaluation so we have to devote to reduce uncertainty of it. Although accuracy of interpretation parameters is regarded as critical factor of uncertainty of petrophysical interpretation result, measurement and interpretation model are not complete also so play important role of uncertainty in reality. Uncertainty analysis may help to find cause of uncertainty but it tend to assume perfection of interpretation method. In this paper, examples of problem of current measurement tools, flaw of interpretation and uncertainty analysis method are showed, and importance of understanding of cause of uncertainty, and validation with other information like oil/gas shows, are stressed to reduce the petrophysical interpretation uncertainty.

This paper attempts to review effective methodology for evaluating geologic risks in the distributions and properties of petroleum system-related sediments, mainly focusing on reservoir/source rocks and migration carrier beds. Since geologic uncertainty may differ depending on the data conditions in the prospect area or targeted basin, appropriate sedimentologic and geologic methodology for predicting the distributions and properties should be selected for better risk assessment in each data condition case. In case the data are sparse or spatially limited in and around the prospect area, it can be effective to apply the methods examining the relative position witin the basin, sediment supply systems, and depositional systems, with practical use of present analogues, the source-to-sink theory, and numerical modeling. If three-dimensional sesimic survey data are available, then effective information can directly be obtained for risk assessment. In case the data are dense in the prospect area, the integrated evaluation of geophysical data, deterministic modeling and geostatistical modeling can be applicable for high-quality and quantitative risk assessment.In this case, deterministic geologic and sedimentologic data should be used as geologic constraints in the geostatistical modeling.

Even in the sparse data case, it can be suggested that multiple combination of distribution/peroperty analysis methods provides key information for geologic uncertainty reduction and accurate risk assessment. Improvement of data density and quality can be another key point for precise risk assessment.

This paper will focus on the development in assessment methodology over the last decades, why the industry is updating their methodology, and the added value this gives the assessors and the decision makers. Improved resolution in prospect definition, from 2D to 3D seismic data, yielded multiple reservoir prospects, which necessitated an update of the risk and resource assessment methodology. In most cases, detailed prospect models with separate closures and reservoirs will have multiple success scenarios. For correct statistical modelling, risk and volume dependencies as well as possible fluid communications between the reservoirs must be honoured.

When 2D seismic data dominated in exploration, resolution was low, and most prospects were modelled as single containers. The increased use of 3D seismic data in exploration resulted in subsurface resolution improvements with multiple compartments and multiple reservoir prospects. Therefore, prospect assessment methods must evolve and enable us to create correct probabilistic representations of such detailed geo-models. Today, the objective of prospect assessment is not only to model the overall chance of success （COS）. It is also critical to model the COS of the different success scenarios and the volume distributions, given that all reservoirs/compartments are tested. To capture all prospect success scenarios in a probabilistic model, it is imperative to handle parameter correlations, risk dependencies, and potential ?uid communications between the different reservoir units/compartments in a prospect. The objective is to enable geoscientists to recreate their understanding of the uncertainty, risk, and possible success case scenarios of prospects in a stochastic model. The established industry best practise for prospect assessment is a four-step process:

A） First, defined the possible success and failure combinations of the prospect, given that all reservoirs/compartments are tested.

B） Break the prospect down into individual reservoirs/compartments that can succeed/fail independently; these are assessed individually with respect to risk and volume.

C） Enrol the assessed reservoirs/compartments into a prospect where risk dependency and possible fluid communication are modelled.

D） Sort success trials from the prospect simulation into groups representing the different possible success scenarios identified in the prospect description.This article will show four different assessments of a multicompartment prospect to illustrate the differences in results going from a simplistic to a geologically realistic assessment.

In addition to geologically realistic assessments of the individual prospects, consistent and comparable analysis is a prerequisite for building and maintaining an exploration portfolio that delivers the predicted results. The last part of this paper will discuss how pre- and post-well analysis and portfolio performance tracking can be used as tools to improve and maintain the exploration performance.

We revisit JOGMECʼs prospect evaluation strategy based on post-well audits for exploration wells from 2007 to 2015. Post-drill statistical results demonstrated relatively good match with predrill geological chance of success and reserves estimation. However, predrill predictions might not sufficiently differentiate high- and low-risk prospects, and large and small reserves. We identified key technical problems to further improve geological risk evaluation. The key elements include appropriate evaluations of DHI （Direct Hydrocarbon Indicator）, petroleum system, and rock volumes.

Since our targets will be geologically more difficult with maturity of oil and gas exploration, it is important to pursue key technologies and to improve risk/volume evaluation for future exploration performance.

More than 30 years has passed since quantitative and probabilistic geological risk analysis was introduced to the oil and gas exploration industries, which is a method composed of two processes 1） to get Chance of Success （CoS） by multiplying grouped geological factors （trap, reservoir, source rock and so on） that are essential for hydrocarbon accumulation in the prospect and 2） to get probability density function （PDF） of resources by running Monte Carlo simulation.

However, it is difficult for some prospects to apply above steps simply due to their geological complexities, which include multi-reservoir, multi-segment and multi-scenario.

Multi-reservoir prospects have more than one target reservoir. In this case, CoS and PDF for resources should be calculated for each reservoir at first, then combined CoS （probability of at least one reservoir has hydrocarbon accumulation） and combined PDF （weighted summation of PDFs for all cases for single or multiple reservoir discoveries with relative probabilities）. Dependencies of geological risk factors and resources calculation parameters between target reservoirs have to be considered.

Multi-segment prospects are separated to more than one segment by faults or other geological obstructions, where combined CoS and PDF for resources can be calculated like as multi-reservoir prospects.

In multi-scenario prospects, there are more than one aspect （scenario） for any of geological factors required for hydrocarbon accumulation. For example, some 4-way dipping anticlinal traps becomes fault-dependent 3-way dipping traps when hydrocarbon-water contacts are deeper than specified depth. In such prospects, probability that the fault seal works can give relative probabilities for both scenarios and separated PDFs for both scenarios can be combined by using those relative probabilities.

Reasonable geological risk and uncertainty models can be constructed for most of prospects in oil and gas exploration businesses with careful combination of above three types of geological complexities.

It is generally said that E&P business have high uncertainty. Although geological risk and reserves risk are major risk for the business, the other risk factors such as resource price, governmental approvals, operatorʼs financial condition has been more noticeably attributed to the actual commercial non-success. Under the circumstances, JOGMEC has made a trial effort to work on project evaluation with a risk-based approach. Risk factors for E&P business are as follows. 1） Underground Resource Risk （Discovery Risk, Reserves Risk, Productivity Risk） 2） Operation risk （Cost Increasing Risk, Schedule Delay Risk, Safety and Health / Environment （HSE） Risk） 3） Management Risk （Team/Human resource risk, Operatorʼs Financial Risk, Strategic Risk） 4） Marketing Risk 5） Financial Risk （Financing Risk, Interest Rate Risk, Liquidity Risk） 6） Country Risk （Macro-Risk （Politics and Economics）, Legal Taxation Risk, Currency Exchange Risk, Nationalism Over Natural Resource, Changes in legal and contract term） 7） External Factor Risk （Price, etc.）On the project evaluating stage, it is necessary to grasp the above risks as precisely as possible, then confirm the control measures for each risk factor, and evaluate whether the control measures would make the risk to mitigate to a reasonable level. JOGMEC has been working on quantification of geological risk / reserves risk and on economic evaluation taking stochastic method into account since the agency established. Besides, in order to respond to the need for strengthening governance in the recent financial business sector in JOGMEC, the above mentioned risk-based approach has conducted on a trial basis.

Cenozoic carbonate buildup reservoirs having a vuggy pore system often show complicated reservoir quality and distribution due to near-surface diagenetic modification including subaerial exposure events. In this study, we evaluate the effects of near-surface diagenesis on pore systems of the Daito Formation carbonates on Minami-daito Island as an analog for carbonate buildup reservoirs. Based on lithological association in a framework of uplifted coral reefs, the outer rim zone and inner land area of the island correspond to the reef margin and platform interior, respectively.

Petrological study revealed that dolostones of the Daito Formation, in which dolomite typically occur as cements and fabric preserving replacements, have a similar pore system to limestones. Dolomitization in this formation is observed mainly on the outer rim zone and in the eastern part of inner land area. Dolomite distribution is controlled by reef topography rather than facies distribution.

Porosity - permeability of this formation is strongly influenced by dissolution pores such as moldic pores and vuggy pores. The carbonates are divided into 4 rock types （RTs） based on pore characteristics （RT 0; cemented rock, RT 1; isolated moldic pores dominant, RT 2; dissolution pores of corals dominant and RT 3; connected vuggy pores dominant）.

RT 0 and RT 1, which are composed of isolated pores, generally show permeability lower than 3 mD regardless of porosity values. Because RT 2 and RT 3 have a well-connected pore system, these are plotted on moderate to excellent porosity - permeability area.

Reservoir quality and its distribution are controlled by multiple near-surface diagenetic events related with the reef topography, facies and dolomitization. Calcite/dolomite cementations have been developed on the outer rim zone and formed the RT 0 and RT 1 rocks. In contrast, dissolution dominated in the inner land area and formed RT 2 and RT 3.

In recent years, owing to the advances in technological development, research and development of unconventional energy resources have been activated. In particular, development of shale gas and tight oil in the United States will have a great influence on the energy supply in the world in the future. Interest in tight oil is increasing in China because China has fascinating unconventional energy resources, and among them, tight oil has the potential to become a promising energy resource. In China, it is expected that reservoir characteristics and availability of tight oil are clarified with the progress of the geological survey, and that the development of tight oil is advanced. In this paper, we examine the availability and potential capacity of tight oil resources in China. On the basis of the characteristics of tight oil reservoir in the United States, the situation and potential of Chinaʼs tight oil resources are examined through the comparison of the geological availability environment of tight oil reservoirs in China and the United States.

At a Steam Assisted Gravity Drainage （SAGD） project on Japan Canada Oil Sands Limited （JACOS）, CANADA, steam leakage from a wellhead of a steam injector was discovered. The steam injection was ceased to eliminate the effects on the environment. Some damages on the wellhead were suspected as the cause of the leakage. In order to inspect an inside of the wellhead due to steam leakage, a “Freeze Application （Ice Plug）” without killing the steam injector was safely applied to remove the wellhead. The integrity of the ice plugs was confirmed by pressure tests, and then steam leakage points at the wellhead were safely remedied with minimal fluid loss to the formation. A typical “Freeze Application” process is shown as below. 1） An area is displaced with a “plug” of uncontaminated bentonite gel mixed to thick slurry. 2） The gel plug is frozen in place by use of dry ice （solid form of carbon dioxide） contained in a cribbing and tamped manually to keep in contact with the area to be frozen at all the times. 3） Pressure testing is always conducted to one and a half times the surface pressure of the well, or to the maximum wellhead pressure rating. In this paper, it consists of background of the steam leakage from the objective steam injector, investigation of remedial action, and then selection and introduction of “Freeze Application”.