Wednesday, October 04
The oil and gas field operators use different techniques to perform production forecasting, and depending on the objective, the process complexity varies. These forecasts can be for a couple of months or for numerous years and the scope of the process can target a well, a reservoir, a field, a country, or even the global assets of an organisation.
Integrated production capacity planning is the process that assimilates production system inefficiencies, and helps to identify, manage and eliminate the limiting factors on the production supply and value chain. The process also guides the field operators to make informed investment decisions to design the optimum system capacity that can deliver the committed hydrocarbon production targets for a given set of constraints.
This session will focus on how the Integrated Asset Models (IAM) can be utilised for defining the optimal system capacity by implicitly sensitising the system parameters and their uncertainties; hence preventing the loss of the economic potential of the asset.
Data-driven analytics and forecasting becomes more and more popular in the industry especially in brown field operation and development. Many oil and gas companies today invest in advanced data analytics and data mining to identify the patterns in behaviour of the wells and equipment.
This session will encourage discussions around four topics:
- What's New in Data-Driven for Production Forecasting?
- Advanced Analytics, Artificial Intelligence and Data Mining
- Simulation vs. Data-Driven Forecasting
- Data Quality and Integration Challenges and Impact on Production Forecasting
We will begin with the discussion around existing and new methodologies to forecast production based on data. The participants are encouraged to share their knowledge and experience using new techniques for forecasting.
The participants will then share their experience in using advanced analytics, AI and data mining for forecasting of production of wells. The discussion will be around data requirements, benefits, challenges, sharing experience and tips, limitations and considerations.
We will then follow with the discussion and comparison between forecasting based on reservoir simulation versus purely data-driven forecasting. The participants are encouraged to share their experience doing different types of forecasting for different types of reservoirs.
After that, the participants will discuss the data quality as it is the key aspect in any type of analysis. The session will conclude with the discussion around the data integration the challenges that it can introduce when it comes to forecasting.
The oil price situation has shifted oil and gas companies’ focus from maximising production to maximising return on investment, which calls for higher levels of integration between technical and commercial disciplines.
Production forecasting is not only used for field development planning but also in support of a number of business decisions such as portfolio management, acquisitions and divestitures, listing, public reporting or reserves based landing. Although the techniques used to generate production forecasts in these contexts are similar to the ones used by asset teams for field development planning (analogies, decline curve analysis, material balance or simulation), the approach generally requires further alignments with overall business strategy, financial objectives, and accelerated decision cycles.
This session will explore how further integration between technical disciplines and petroleum economics influences production forecasting considerations to maximise profit, how production forecasting is conducted in other contexts than field development planning, and how widely recognised industry guidelines (such as the SPE Petroleum Resource Management System) define production forecasting best practices and economic limit tests.
Production forecasting takes place at many different levels in organisation and for many different purposes such as reservoir management and field development to meet the future demands. The challenge of workflow orchestration has largely been unresolved in forecasting due to integration of various activities from subsurface to integrated activity planning. The numerical, analytical, IAM based, and data-driven methods are efficiently need to be selected based on triggers and type of forecasting. Automating the routine work involved in setting up the data that is input to the forecasting tool, configuring various model parameters, invoking the tool, analysing the output, and generating the desired reports and graphs. Forecast accuracy and reliability are the key measures in the business process.
Since the automation is increasing more and more in digital oil field, this session mainly covers developments and advancements made in the area of automating the forecasting workflow and its business process for improving efficiency and reliability of results.
Thursday, October 05
Forecasting techniques used in conventional reservoir have not proved to be effective for unconventional resource reservoirs. This becomes quite evident when attempting to apply conventional decline curve analysis techniques. Due to the unique physics and geology of unconventional reservoirs, decline rates change with time usually three to four times over the life of the well. Beginning with dramatic declines of 65–80% during the first year of production, 35–45% for the second year, 20–30% for year three, and continuing at a constant +/-5% for 20 to 30 years. Therefore operators have mostly employed “Type Curves” (using data mining to assemble just a few years early production from offsets) to predict rate and recovery with a different curve for each unconventional play. Type curves have not always been accurate for forecasting. The alternative has been to build reservoir models and run simulations; however, the data required has not been available as operators save money on data collection. Less than 8% of US unconventional wells have logged or gathered any data along the horizontal lateral. Also, the traditional PTA techniques to determine pressure profiles has been prohibitive due to the extremely low Nano Darcy permeabilities of unconventional reservoirs.
Therefore, forecasting, in particular while setting expectations regarding initial well productivity, defining optimum drilling and completion strategy, determining optimum well spacing, and identifying overall resources requirements has been problematic. Again operators have turned to various alternative methods.
In this session, we will examine production forecasting in unconventional reservoirs, while examining the challenges around subsurface characterisation, well completion and stimulation practices, and integrated operations with field development dynamics. Various forecasting methods are supporting unconventional forecasting: numeric simulation, data mining and time series, or a combination of them.
Production forecasting is an essential part of the development and management of a company’s reserves. A proper and reliable forecast will allow the company to optimise the production potential while maximising the use of their facilities at minimal costs. In this session, we will compare different aspects of production forecasts to further visualise how we can tailor our forecast methodology to suit our business needs.
The top-down versus bottom-up forecast strategy present different approaches to forecasting, with existing and future facility framework, a generic reservoir development programme may be derived from a set of average well rates, to give a top-down forecasting approach. If details of well-by-well performance, vessel-by-vessel capacity, expansion, and maintenance schedule are available, a bottom-up forecast could be generated, thereupon these two methods could complement each other.
Green fields vs. brown fields production forecasting methodology has its own challenges and requires slightly different approaches. For new fields with limited data available for trending; similar fields’ data could be used as analogs to start the forecasting. For older fields, adequate historical data are incorporated into the forecasting process. Depending on the needs of either field development in new fields or field maintenance in old fields or the gain/loss potentials, the forecast must be scrutinised and may require more prudence in its designs.
Transitioning between short-term and long-term forecasting may require different toolkits to allow smooth and logical link between the two timeframes. This smooth transition between these two approaches will produce a more applicable forecast.
A good forecast will allow the company to make wise decisions for further investments to optimise its asset development and business potentials. New approaches or even optimisation of available tools play an important role in generating a forecast which other areas can rely and develop their strategies to suit the company’s needs.
Reservoir uncertainty is one of the biggest challenges in production forecasting activities; in order to have meaningful and reliable predictions, proper focus is required towards risk analysis and uncertainty assessment tasks. In this session, we will discuss recent or innovative practices to address, analyse and assess the most common risks and uncertainties, and its respective impact in production forecast results.
Subtopics to be included as part of this session are:
- Reservoir heterogeneity impact, including fractured reservoirs, carbonate reservoirs, among others
- Risk assessment for production forecasting scenarios
- Field development plans (usually long-term forecasting exercises) and production optimisation (fit-for-purpose methodologies with short-term applicability)
- Probabilistic and deterministic production forecasting; selection of appropriated approach (simple vs. sophisticated) and impact on uncertainty
- Best practices and lessons learnt; pragmatic workflows in lieu of data limitations and big reservoir uncertainty
Production forecasting and its related risk and uncertainty quantifications are one of the key contributors to the field development decisions. However, to date there have been limited standard workflows established in the industry to address these daily challenges. This session will highlight some of the industry’s best practices and technical workflow developments through technical case studies on the field experiences, their successes and failures.
The technical case studies will aim to compare methodologies and discuss their differences, advantages, and disadvantages. Potential topics include:
- Ingredients of successful production forecasting
- Inherent risk and uncertainty of forecasts
- Learnings from other industries
- Benchmarking of different approaches
We will then discuss how these example case studies and their results could lead to a guideline or if it at all standardisation of production forecasting workflows are possible.