First EAGE / TNO Workshop

Modeling of groundwater flow through basins has to be based on regional-scale hydraulic conductivities. We show how regional conductivities can simply be calibrated using direct and indirect measurements of hydraulic heads and fluxes. Groundwater models are conventionally based on recharge flux specified on the model’s top boundary. In the 1960s Tóth presented a model of regional groundwater flow based on head specified on the top boundary. This approach resulted in the discovery of local, intermediate and regional flow systems. On the other hand, a top head condition may lead to unrealistic recharge rates. Combining the advantages of both flux and head boundary conditions, we analyze groundwater flow using two models: a flux model and a head model. The difference between the two solutions shows the regions where measured heads and fluxes can calibrate the initial hydraulic conductivity, as well as the regions where measured heads and fluxes cannot. Using Darcy’s law to the differences results in calibration. This so-called Double Constraint Method can be made more reliable by repeated calibrations under different hydrological conditions. A Kalman Filter then determines an uncertainty that is appreciably smaller than the observation error.

The development of an extensive integrated pressure database since 2002 in combination with integrated approaches to characterize and interpret current basin hydrodynamic systems in the Netherlands have resulted in knowledge and understanding of spatial variations in hydrodynamic and pressure conditions in relation to its geological framework and burial history. This paper discusses integrated application of hydrodynamic-based approaches with petroleum science based modelling approaches to identify reservoirs of lateral hydraulic continuity and vertical leakage zones. The approach will be illustrated with case studies from the Netherlands.

This paper presents preliminary results of the analysis and interpretation of the observed pre-production fluid overpressure distribution in Rotliegend sandstones in block L8 in the Dutch offshore. The general objective of this study is to develop workflows to improve the understanding of timescales of pressure dissipation in relation to regional to basin scale fault models, including fault sealing properties. For this case study we used Eclipse reservoir modeling. The model for the Permian Upper Rotliegend Group includes the numerous gas fields in the area. Based on the configuration of these gas fields in combination with the observed regional overpressure distribution the original fault model could be improved. By creating a ‘dynamically stable’ model, fault multipliers were determined which gave a pressure drop below 1 bar over a period of 5000 years. The resulting distribution of the fault multipliers over the model area clearly identified differences within the study area. A large part of the estimated fault sealing capacity could be explained by the geometry of the fault (offset and sand-sand juxtaposition). The simulations also showed that only a small area of sand-on-sand juxtaposition requires a reduced fault transmissibility in order to retain the observed overpressure for any length of time.

The goal of this study was to find correlation between springs, as natural discharge features and their flow systems by numerical simulation in case of thick carbonate regions. The scenario modelling could prove the existence and operation of gravity-driven flow based on two different scenarios. The differences in the effect of the water table, basin geometry and geology caused different flow pattern. The flow is dominantly lateral in the flow domain. The smoothed differences of the water table could cause mainly local but also intermediate flow systems depending on the depth of the basin. The structures did not caused such a significant effect on flow pattern. However the geometry of the basin and the existence of low permeability strata could efficiently restrict the flow. The folded structure in case of a relatively shallower basin modified the flow pattern. In case of a deeper basin and confined situation complex flow patterns could evolve under confining strata. The position of the erosion basin determines the discharge places at the boundary of unconfined and confined system. The effect of temperature is not so sufficient. Those parts of the basin can preserve heat where the low permeability strata limit the effectiveness of gravity-driven flow.

Understanding and insights into the fluid flow systems as well as heat transport processes of carbonate systems can be significantly improved with numerical models. Numerical simulations have provided new insights into the processes controlling fluid flow and heat transport within the deep and thick carbonate system of the Buda Thermal Karst, which can be considered as a pilot area to examine the importance of these different driving forces. Different model scenarios of four cases were tested to represent snapshots of the fluid evolution of the system, from the fully confined carbonate stage in the Late Miocene through to partly and completely unconfined conditions (over the left Buda block), the latter representing the recent situation. The effect of changes in different parameters of the numerical model on the potential field and temperature distribution were examined. The preliminary results highlight the effects of paleo-recharge and cover formations on heat distribution and dissipation. In most cases, conditions led to strong thermal convection cells which might have played an important role in system evolution. The results should later contribute to a better understanding of hypogene karstification, and utilization of geothermal resources and hydrocarbon resources of deep carbonate systems.

The importance of regional hydrodynamics in the site characterisation workflow for CO2 storage is often underestimated. Examining the reservoir in the larger context of the regional aquifer can provide insight into the short and long term response to injected CO2. The CO2CRC Otway Project located in the onshore Otway Basin, Victoria, was the first CCS demonstration project in Australia (Cook 2014). The CO2CRC undertook a detailed regional, hydrodynamic study of the basin, focused on the target reservoirs, but including evaluation of the overlying aquifers. The hydrodynamic model significantly improved the reservoir simulation for the initial injection and enabled estimation of properties associated with shallower, data poor, secondary storage targets. Groundwater monitoring is considered essential to demonstrate to stakeholders the ongoing integrity of the groundwater in the vicinity of the storage activities. However, simple numerical modelling based on the groundwater aquifers at the Otway site demonstrated that the pressure and chemical signal of buoyancy driven CO2 into this aquifer is only detectable in the immediate vicinity of the entry point. Therefore, the location of groundwater monitoring wells is of critical importance, given that the volume of the aquifer is large and the volume sampled by an individual well is small.

In the Haldimand sector of Gaspé, Québec, Canada, a study was carried out to assess the potential risk on a shallow fractured rock aquifer system due to development of a tight sandstone petroleum reservoir. Petroleum exploration wells are being drilled in the forested core of a hilly 50 km2 peninsula by the sea (up to 200 m amsl) and where local residents rely on groundwater wells for their water supply. The study used existing hydrogeological, geological and petroleum exploration data and more recently acquired field characterization data. Groundwater and surface water sampling within a 2 km radius of a proposed new drill pad. All samples were subject to chemical analyses. Fracturing controls groundwater flow especially in the upper 15 m of the rock aquifer. Recharge occurs on topographic highs where the glacial till cover is thin. Quite wide variations in groundwater geochemistry were encountered. Groundwater residence times can thus be quite long. Methane is of mixed origin but is preferentially associated with the relatively more evolved water types. The SALTFLOW model was used to simulate density-dependent groundwater flow and salt transport within the peninsula as well as the adjacent highlands along a 2D vertical section.

From fundamental hydraulic principles, an equation defining the surface of a tilted OWC was derived. The derived equation can be used directly to delineate the OWC for a given hydrocarbon accumulation; and to map hydrodynamic traps. An easy-to-understand approach to map hydrodynamic traps by employing some simple map operations has been developed. we present a Middle Eastern example to illustrate our developed approach for mapping hydrodynamic traps. The hydrodynamic effect favors the studied region.

Hydrodynamic reservoirs have been widely recognised and described since the 1950s when petroleum hydrodynamics was first described by Hubbert. More recently deep drilling of high pressure sediments in petroliferous basins has revealed hydrodynamic reservoirs, where aquifer drive comes from the association of high pressure sediments surrounding laterally draining reservoirs. Lateral drainage can only be confirmed where direct pore pressure measurements from a series of vertically stacked reservoirs show pressure "reversals" (high overpressure to low overpressure with increasing depth). In many situations there are no direct pressure measurements but pore pressure estimation in associated shales and the log signature of those shales can generate diagnostic criteria for lateral drainage. Conventional, upland-driven hydrodynamic flow leads to reservoir overpressures less than the surrounding sediments in most cases. What we are describing as "unconventional" hydrodynamic flow driven by dewatering of high pressure sediments has the characteristic of low overpressure reservoir in association with high overpressure sediments. Both systems are associated with the tilting of hydrocarbon-water contacts and obey the same rules governing the angle of tilt. However, the timing of fluid release for lateral drainage will influence the trapped fluids and the distribution of hydrocarbons. There is a worldwide dataset of examples.

The concept of hydrodynamics and tilted hydrocarbon-water contacts excites passion but remains controversial. It was recognized in the later 1950s that flow through the aquifer was capable of tilting hydrocarbons held in a trap. Subtle traps resulting from hydrodynamic conditions encountered in petroleum basins were described for many years (e.g. Hubbert, 1940, 1953). The aim of this paper is to show, through the example of the Llanos basin (Colombia), how 3D basin modeling can help in determining the hydrodynamic traps and in evaluating the related volumes of hydrocarbon originally in place.