Ian Lerche

Flooding of lignite mines: isotope variations and processes in a system influenced by saline groundwater.

The quality of both groundwaters and surface waters that arise during flooding of abandoned lignite open pits are influenced by regional and local factors. A typical regional factor is due to oxidised sedimentary sulfides. A more local factor is the interaction of shallow water with highly saline groundwater, which is important in Merseburg-Ost (Germany). Investigation of this system is aided by the use of many environmental isotope tracers but special problems can arise. In order to reveal processes in the mine environment (shallow groundwater, lake water) and to characterise mixtures with saline groundwater results are described using the tracers δD, δ18O, δ13C, δ34S, 87Sr/86Sr, 3H, 14C,39Ar, and 222Rn. Deep highly saline groundwater had a radiocarbon concentration typically below 10 pMC. The values of δ13C(DIC) are around−5 ‰. As δ13C of the aquifer rock samples (Permian, Zechstein carbonates) was in the range of−6···+5 ‰, residence time corrections based on δ13C are questionable. Additional checks with 39Ar, as well as results from the variationof δ18O (or δD) with respect to the salinity, emphasise a Holocene age; as is also the case for most mineralised groundwaters and also for water having a low δ18O (and δD). For saline groundwater residing in the Zechstein aquifer the measured δ34S values of about 12 ‰ are close to those expected from the literature. In contrast, the 87Sr/86Sr ratio of dissolved strontium is far from the values anticipated for the aquifer rocks despite there being proportionality between the chloride concentration and the strontium concentration. Furthermore, the proportionality is not valid in lower mineralised water. The 87Sr/86Sr ratio can, therefore, hardly be used as a tracer for the distribution of ascending saline water. The amount of salt-water coming from below into the residual quarry basins is an essential contribution to the lake inventories. Therefore, 222Rn was used to assist in determining the renewal of salt-water layers that formed in deep lake locations. In the deep zones 222Rn concentrations up to 6 Bq/l were measured but were dominantly in equilibrium with 226Ra, which was found in all higher mineralised groundwater samples. Excess radon was limited to just a few decimetres above the lake sediment surface but does not appear to be caused by continuous groundwater discharge. Hydrochemical investigations of groundwater from the Quaternary aquifer were carried out over the last six years before flooding was complete. Apart from a slight downward shift of the average sulfate concentration, other changes showed virtually no trends. An increase of the sulfate concentration was mostly correlated with a decrease of δ34S for individual sites only, but not for the whole ensemble of sampling locations. Sulfate from pyrite oxidation plays an important role but cannot be attributed unequivocally to coal mining. There are hints that the conditions closer to the basin edges may differ from those remoter parts of the flood plain.

Economic Exploration Assessment of Hydrocarbon Leakage Across a Fault. II. Charged Sand Before Faulting

That faults cross-cut sands and cause leakage of hydrocarbons, either out of the upthrown and downthrown sand system or from the downthrown sand to the upthrown sand, is well known The decision to drill either or both of such sands depends on what it is one assesses as the likely hydrocarbon potential of the total system and also for each sand independently. This paper provides a quantitative procedure to determine a strategy for deciding what to drill. The economic worth for such a sand system depends on the chances the sands were originally hydrocarbon charged, and on the leakage possibilities for each sand, together with the recharge of the upthrown sand from the downthrown sand, and seems not to have received a quantitative investigation to date. In addition, at the exploration stage the amount of detailed information about such a fault cut system is limited, so that any and all assessments of parameters influencing the economic worth are also uncertain. This paper shows how one can include uncertainties caused by such “fuzzy” parameters in attempting to decide on a drilling strategy geared to satisfying corporate mandated economic returns criteria. Numerical illustrations are given to show how the strategies to decide to drill either only the downthrown sand, or only the upthrown sand, or to drill both sands can be brought to a rational form so that corporate decision-makers can compare and contrast the three strategies under the same quantitative framework.

Potential Leaks in Hydrocarbon Reservoirs: Impact on Resource Assessments:

Evaluation of the likely resources to be found in a four-way structural closure is undertaken when there is a chance of an open fault intersecting the structure that can leak hydrocarbons. Because one does not know whether the putative hydrocarbons exist in the structure and, even if they exist, whether they fill the structure to the point that leakage can occur, such assessments have, of necessity, to be probabilistic. We show here that there are two main contributions to the uncertainty on the expected resources: the first arises due to the uncertainty of which outcome will be present, and the second arises because the parameters entering the resource assessment each have uncertainties. Both types of uncertainty components have to be considered when evaluating the expected value of the resources and its total uncertainty. The probability one will find resources in excess of a particular value is also evaluated including both components of uncertainty. The dominance of particular parameters to the uncertainty is also evaluated. Simple numerical illustrations show how one uses the procedure in practical situations to evaluate the individual components and also to address properly the worth of obtaining further information, and the type of information, to narrow the range of uncertainty caused by the probable presence of the leaking fault and the amount of hydrocarbons it might be able to leak from the structure.

Leakage along faults : Capture by an overlying sand

This paper considers the capability of a sand, cut by a fault, to capture hydrocarbons supplied from below along a fault. The volume supplied, the capacity of each part of the sand, and the fractions of volume supplied that could be diverted to each sand, are all significant influences on the amounts of hydrocarbons that could be captured. Simple deterministic examples are given to illustrate the patterns of response that ensue when individual parameters are varied, including the amount of hydrocarbons lost from the system. In addition, when all of the parameters can be uncertain, as is most usually the situation when undertaking hydrocarbon exploration estimates of possible reserves, Monte Carlo calculations are used to show which of the uncertain parameters is causing the largest fraction of uncertainty on the fill and loss amounts from the system. In this way one can determine where to expend effort to narrow down the range of uncertainty of the parameters causing the dominant contributions to fill and loss uncertainty, with out having to spend inordinate amounts of time and effort in attempts to determine parameter ranges better for those parameters that hardly influence the uncertainty on hydrocarbon fill and loss.

Economic Exploration Assessment of Hydrocarbon Leakage across a Fault. III. Hydrocarbon Leaks and Fault Locations

This paper addresses the concerns of potential hydrocarbon leakage through a fault when there are changes in the location of the fault, its orientation, and its length in relation to a given sand body of known width. The probability that hydrocarbons interact with the fault is expressed in terms of the overlap length of the fault in the sand, and in terms of the sand width as well. When the probability that the fault is open or shut to hydrocarbon flow is added to the geometric probabilities as a basic variable, the uncertainties of the five parameters (fault length, sand body width, fault orientation, fault location in relation to the sand body, and probability the fault is open to hydrocarbon escape) are addressed in terms of Monte Carlo representations of the chances of retaining or losing hydrocarbons. The dominant contributions to the uncertainties of the fraction of hydrocarbons most likely lost or retained are also given so that one can determine which of the five basic parameters need to have their ranges of uncertainty narrowed first in order to improve on the probable estimates of hydrocarbon retention. Numerical illustrations are given to show how one goes about assessing these factors for a sand cut by a fault, and to illustrate the chances that the sand is hydrocarbon bearing or not due to possible loss through the fault.

Oil and Gas Estimates for Arctic National Wildlife Refuge Area 1002, Alaska

With minimal information from 5 seismic lines, and 7 projected pseudo-wells from the Arctic National Wildlife Refuge (ANWR) Area 1002 (undeformed area Brookian Sequence), Alaska, USA, state of the art basin modeling software was used to provide quantitative 1D and 2D patterns of the basin evolution in terms of burial history, hydrocarbon generation, migration and accumulation. With the available geological information a stratigraphic profile was constructed containing 10 layers representing different depths and formation ages. From the burial history analyses of the 7 pseudo-wells, forecasts were created at different spatial-temporal slices of cross-sections and contour maps for different variables, including TTI, kerogen fraction type II, fluid flow velocity, overpressure, and oil and gas available charges. This paper illustrates how a sensitivity analysis study provides information about which geological parameters involved in the system are causing the major contributions. Relative importance, relative contribution and relative sensitivity are examined to illustrate when individual parameters need to have their ranges of uncertainty narrowed in order to reduce the range of uncertainty of particular outputs. The Monte Carlo simulation procedures using Crystal Ball® as an interface for risk analysis are fast, taking on average 10 minutes on a laptop computer to perform 1000 iterations with 236 uncertain variables. Different groups of runs were performed individually, as well as combinations of different variables according to the module in use (geohistory, thermal history, geochemistry, fluid flow, and oil and gas generation, migration and accumulation) for uniform stochastic distributions in order to identify which groups of variables were causing the largest uncertainties to hydrocarbon charge estimates. Results indicate about 2.3 Bbbl as the maximum oil available charge. The ranges of uncertainty for different parameters were modified from their nominal values (for instance: the uncertainty in the oil fraction produced by Type II kerogen was modified from 50% to 25%; the geothermal gradient uncertainty at 23.8 MyBP was modified from 50% to 30%; migration loss uncertainty for formations deposited approximately 33 to 49 MyBP was reduced from 50% to 25%; and the uncertainty on the amount of kerogen type II for the same layers was lowered from 50% to 25%) emphasizing that the maximum contributions to the uncertainty for oil accumulation estimates were for kerogen type II (about 13% contribution) and the oil fraction produced from kerogen type II (12% contribution) for formations currently between 3.4 and 4.6 km depth and deposited approximately 33 to 49 MyBP. Just these two major contributors provide about 25% of the total uncertainty for available oil charge. For gas the maximum charge available reached was about 46 Bm3. Lowering the uncertainty of parameters (such as the oil fraction produced from type II kerogen from 50% to 25%, the geothermal gradient uncertainty at 33 MyBP from 50% to 30%, the geothermal gradient uncertainty at 28.5 MyBP from 50% to 30%, and the uncertainty on the kerogen type II amount at layer 10 from 50% to 20%) allowed one to determine that the major contributors to the uncertainty were from the uncertainty of the oil fraction produced by type II kerogen (about 12%) and from the uncertainty on the fraction of type II kerogen (11%), occurring in formations currently between 5 and 5.1 km depth and deposited between approximately 71 and 112 MyBP. The total uncertainty for gas charge available from these two major contributors sums to about 23%. The ranges in estimates of oil and gas charges available today are in contrast with previous assessments due to the inclusion of the influence of dynamic range uncertainties for the parameters.

A Binomial Approach to Selecting Optimum Working Interest: Corporate Risk and Corporate Confidence:

Current methods for determining the optimum working interest (OWI) one should take in an exploration project are beholden to underlying basic utility theory. As such they suffer from the generic consequences that: the OWI is never guaranteed to be always less than unity as it must be; that the OWI does not always increase as the net present value of a project is increased as it should; and that the OWI is automatically set to zero when the expected value of the project is less than zero, despite there being an uncertainty on the expected value that could still allow a profitable outcome. As a consequence of these drawbacks, here we set up a binomial procedure for ascertaining the OWI, which method does not suffer from any of the utility-based problems. Several illustrations indicate how the binomial enhancement operates in reality. In addition, one can also compute the corporate confidence in a project using this new approach, and determine the corresponding apparent confidence of any utility-based procedure for determining OWI. This binomial procedure for determining OWI and corporate confidence is, therefore, a reasonable alternative over previously available methods.

Potential Leaks in Hydrocarbon Reservoirs: Costs, Profits and Uncertainties

A potentially leaking reservoir has an uncertainty on the volumetric resources it might contain because one does not know the extent to which it might have leaked nor does one know that it actually does contain hydrocarbons ahead of drilling. The economic aspects of such an exploration opportunity are investigated in this paper, where it is shown that not only do the uncertain volumetric aspects play a role in determining potential profitability but, and often more importantly, so do the potential drilling costs and the future selling price of product, both of which are also uncertain. It is the combined interaction of both the volumetric uncertainties and the economic uncertainties that determine the possible worth of the opportunity. Here we show how to determine the parameters that are most important in contributing to the uncertainty on the potential worth. We also show how to estimate the probability of achieving break-even and how to determine which parameters are causing the greatest uncertainty on the break-even probability. In this way decision-makers in a corporation can determine more effectively where to concentrate their efforts to improve the uncertainty of critical parameters, and they also know which are the critical parameters to be addressed. Time, effort, and money are thus spared by not focusing on parameters that play more minor roles in the assessment of a potentially leaking reservoir as an exploration opportunity.

Conditional drilling of a sand cut by a fault

This paper discusses the Bayesian updating of the chances of hydrocarbon success in a sand cut by a fault when the sand is charged after fault time. Prior to actual drilling the decision to drill either the downthrown sand component or the upthrown sand component depends, in particular, on whether the fault is considered to be leaking a lot or a little and whether any leaked hydrocarbons are then trapped in the upthrown sand. Based on prior-to-drill estimates of parameters one can then determine which is the better probability to drill. After drilling one knows with certainty the worth obtained in one component of the fault cut sand. This information can then be used to update the probability that the fault is leaking a lot or only a little, as well as to change the assessment made prior to drilling concerning the probability that a hydrocarbon charge indeed took place in the lower sand after fault time. These updated parameter values can then be used to re-assess the location off the next drill site in order to improve the chances of commercial success. This group of problems is addressed in this paper with numerical examples to show how the procedures operate in practise.

How Many Monte Carlo Simulations Does One Need to Do? II. Lognormal, Binomial, Cauchy, and Exponential Distributions

This article derives an estimation procedure to evaluate how many Monte Carlo realisations need to be done in order to achieve prescribed accuracies in the estimated mean value and also in the cumulative probabilities of achieving values greater than, or less than, a particular value as the chosen particular value is allowed to vary. In addition, by inverting the argument and asking what the accuracies are that result for a prescribed number of Monte Carlo realisations, one can assess the computer time that would be involved should one choose to carry out the Monte Carlo realisations. The arguments and numerical illustrations are carried though in detail for the four distributions of lognormal, binomial, Cauchy, and exponential. The procedure is valid for any choice of distribution function. The general method given in Lerche and Mudford (2005) is not merely a coincidence owing to the nature of the Gaussian distribution but is of universal validity. This article provides (in the Appendices) the general procedure for obtaining equivalent results for any distribution and shows quantitatively how the procedure operates for the four specific distributions. The methodology is therefore available for any choice of probability distribution function. Some distributions have more than two parameters that are needed to define precisely the distribution. Estimates of mean value and standard error around the mean only allow determination of two parameters for each distribution. Thus any distribution with more than two parameters has degrees of freedom that either have to be constrained from other information or that are unknown and so can be freely specified. That fluidity in such distributions allows a similar fluidity in the estimates of the number of Monte Carlo realisations needed to achieve prescribed accuracies as well as providing fluidity in the estimates of achievable accuracy for a prescribed number of Monte Carlo realisations. Without some way to control the free parameters in such distributions one will, presumably, always have such dynamic uncertainties. Even when the free parameters are known precisely, there is still considerable uncertainty in determining the number of Monte Carlo realisations needed to achieve prescribed accuracies, and in the accuracies achievable with a prescribed number of Monte Carol realisations because of the different functional forms of probability distribution that can be invoked from which one chooses the Monte Carlo realisations. Without knowledge of the underlying distribution functions that are appropriate to use for a given problem, presumably the choices one makes for numerical implementation of the basic logic procedure will bias the estimates of achievable accuracy and estimated number of Monte Carlo realisations one should undertake. The cautionary note, which is the main point of this article, and which is exhibited sharply with numerical illustrations, is that one must clearly specify precisely what distributions one is using and precisely what free parameter values one has chosen (and why the choices were made) in assessing the accuracy achievable and the number of Monte Carlo realisations needed with such choices. Without such available information it is not a very useful exercise to undertake Monte Carlo realisations because other investigations, using other distributions and with other values of available free parameters, will arrive at very different conclusions.