SPE logo

Production Forecasting

12 – 14 May 2014

Dubai, UAE | The Address Dubai Marina Hotel

Technical Agenda

Session 1: Data Gathering Quality Assurance and the Impact on Production Forecasting

Session Chairs: Sarika Verma, Harriot Watt University; Tony Thomas, ADCO

To improve the reliability of reservoir performance predictions, data uncertainties must be accounted for in the predictions. This session will address some of the common issues with data uncertainties and how they impact the production forecasts. It will outline some of the techniques towards assessing the uncertainties and risks related to the data and how to mitigate those risks while forecasting. Sustainable data gathering and quality assurance strategies that are employed towards improved production forecasts and hence better reservoir management will be discussed.

Session 2: Tactical vs. Strategic Decisions: Near Term vs. Long Term Forecasting

Session Chairs: Arash Shadravan, Baker Hughes; Kaveh Dehghani, Chevron 

The main purpose of forecasting is to help decision makers with an informed view of the range of possible outcomes and risks associated with their decisions. These decisions could be strategic addressing current long term asset development plans and prioritisation of a queue of opportunities that may be developed over the life of the asset, or tactical to address day to day production optimisation practices. The decisions could range from reserves estimation and business plan to maximise it, to a simple short term exercise in identifying an optimised location for the next well drilled. This session will be a platform to discuss and present case studies on:

  • The forecasting practices that could address both NOC and IOC views of the region specific strategic decisions and their differences.
  • The appropriate level of rigor to be practiced for the forecasting of different day to day tactical decisions.

Session 3: Fit-For-Purpose Forecasting

Session Chairs: Hamed Wahaibi, Petroleum Development Oman; Olalekan Kayode Fawumi, ExxonMobil

The purpose of this session is to address “fit-for-purpose” production forecasting methodologies and their limitation. This approach implies working from top to down by defining decision targets and uncertainties which gauge the precision level and time required for forecasting and hence the method.  The available production forecasting methods vary from analytical, volumetric, and numerical techniques. At the end of the session, the participant will aim to outline some business scenarios/examples where particular methods were used as “fit-for-purpose”.

Session 4: Probabilistic and Deterministic Production Forecasting

Session Chairs: Jonathan Hastings, Maersk Oil; Yomi Adesimi, Suncor Energy

One of the key choices we make before we begin forecasting is whether to use one or more deterministic forecasts or to take a probabilistic approach. The software, workflows, and even computing power are now at a point where sophisticated probabilistic evaluations of subsurface uncertainty are simple and routine, yet the deterministic forecast is not dead! This session will look at the current practices of both, their pros and cons and the common pitfalls. In particular:

  • Common probabilistic misconceptions and, what, if anything, is a P50 forecast?
  • When is a deterministic forecast more appropriate?
  • How can we properly define and quantify all the relevant uncertainties in a probabilistic analysis?
  • Do we have the systems in place in order to make decisions on the basis of probabilistic forecasts?

Session 5: How to Improve Forecast Quality

Session Chairs: Hugo van Rossem, Shell; Charles-Henri Koeck, IFP Middle East Consulting 

On the one hand, we generate long term recovery forecasts that meet regulators requirements for resource volume estimates. On the other hand, we make every effort to best inform our stakeholders what the expected or potential production levels will be for the immediate future. Are these objectives actually the same? Are the uncertainties not very different? Do we apply the same methods and QA/QC standards for both types of forecasts? And what are these quality standards? In other words: how can one judge the quality of a forecast? What can we do to improve the quality of subsequent forecasts? Can we learn from other industries? Do we ever look back effectively and learn from “poor” forecasting? If so, how?

In this session, we hope to hear from companies and what their practices are to assure the quality of their forecasts. In particular:

  • What is a “good” forecast? What criteria should we use to determine the quality of a forecast?
  • How do we learn from “bad” forecasts? What methods are best used to ensure we improve the quality of subsequent forecasts?
  • How can we best express and manage the inherent uncertainty in forecasts? Is precision better than accuracy?
  • What can we learn from other industries?

Session 6: Forecasting in the Age of the Intelligent Well and the Digital Oilfield

Session Chairs: Nelson Bolanos, Schlumberger; Ralf Schulze-Riegert, SPT Group

Intelligent completions enable operators to monitor and control individual zones/sections and optimise production or injection programmes. Digital oilfield, facilitated by modern information and communication technology, improves operational effectiveness and overall asset management. Both intelligent completions and digital oilfield provide us with rich data, but it still remains a challenge to effectively utilise the information for the decision making process. We will discuss in this session how production forecasting techniques can help this process.

Potential topics include:

  • What is the data available from intelligent completions and digital oilfield programme?
  • What are the purposes of production forecasting for intelligent completions and digital oilfield?
  • What are the production forecasting methods and/or software tools typically deployed for these scenarios?

Session 7: Green Fields vs. Brown Fields: Mega Fields and EOR Projects

Session Chairs: Christiaan van der Harst, Shell; Hashem Monfared, Schlumberger

Uncertainty quantification and production forecasting for green fields seems to be as challenging as for brownfields. While the former suffer from lack of data and limited production history, the challenge for the latter is the integration of large volumes of data at different scales as well as long production history into predictive tools by using either analytical or numerical techniques. Predictions under uncertainty workflows are yet to address all aspects of practical challenges and this becomes even more complicated with EOR projects, especially for giant fields. The aim of this session is to review the current production forecasting practices for green and brown fields. Deterministic as well as stochastic approaches and the latest workflows will be addressed.

Session 8: Business and Investment Related Forecasting

Session Chairs: Djamel Eddine Ouzzane, BP; Venkatesan Ganesan, Deloitte Corporate Finance

Whilst subsurface uncertainties cast an enormous challenge during production forecasting, issues around business climate and investment related aspects in forecasting are not anything less challenging. External environment challenges are endless which include inflation on capex and opex costs, product prices, taxation, expected return on investments to mention a few. These aspects will have a major bearing on the project feasibility and decision making. This session will focus on external environmental challenges and how the industry and financial experts take-on those challenges. The speakers would offer their insights into modelling techniques generally followed, data sources, and how to sense-check/reconcile modeling assumptions.