The importance of accurate reservoir description is illustrated using reservoirstudies of various pools in Alberta. Special emphasis is placed on reservoirdescription for mathematical models. Studies show that deficiencies in certainbasic data, often considered to be of minor importance, can lead to significanterrors in predicted performance.
Inadequate reservoir description considered in this paper includes errors thatmay occur in determining such basic data as porosity and permeability. Theeffects of performing special core tests on weathered core are also considered. Primary depletion performance in the Pembina Cardium Pool illustrates theimportance of using the correct permeability cut-off to determine effective ornet pay. The significance of reservoir configuration and well location areillustrated by studies of Keg River pools from the Zarna Virgo fields. Waterconing in the Harmattan Elkton gas pool is used to demonstrate how formationdamage can affect predictions of future water production.
Early reservoir studies were generally quite straight forward and relativelyuncomplicated. Primary depletion calculations were based on a Muskat or similaranalysis and waterflooding was evaluated by one of the classic methods such as Stiles or Dykstra and Parsons and later by methods proposed by Craig, Geffen & Morse and Pratts et al etc. Spacing was usually pre-determined andstandard injection patterns were used. With these conventional studies therewas not the need for the same degree of accuracy in reservoir descripti0n asthere is in some of the more sophisticated studies of today.
Today the engineer requires more information from a reservoir study;information that will not only define the reserves and economics of variousdepletion schemes but also assist in optimizing the operations of thereservoir. He needs more information to define the optimum spacing andproduction and injection patterns and how best to complete wells. The latter isespecially true when coning from a gas cap or underlying water isconsidered.
To assist in the evaluation of the more complex reservoir problems thereservoir engineer of today has many more tools at his disposal. One of themost important of these is the mathematical model which is used to simulatereservoir and well performance. The advent of high speed digital computers hasmade this numerical technique economically feasible for most engineeringapplications. Unfortunately, when modelling was first introduced it wasoverrated and hailed as the answer to practically all engineering problems eventhough the initial programs were expensive and often introduced large errors inthe mathematical solutions.
However, even at this time erroneous conclusions obtained from model studieswere often the result of inadequate reservoir description. Basic engineeringprinciples were forgotten in the desire to make use of this latest reservoirengineering tool. It was treated like a magic box and in many cases thesolutions obtained were not studied to evaluate the possibility that they couldbe erroneous. The result of this oversell is that engineers became disenchantedwith the high costs and sometimes meaningless output of some of the earliersimulation studies.