08/01/2007 The recipe for good forecasting is straightforward enough. Enter qualitycontrolled data in simple but well-defined reporting systems. Use advanced technical aids, and blend with a portion of historical understanding and another portion of the ability to predict possible futures. Mix well in a big portfolio assessment and test the results. And learn from your mistakes.
Creating good predictions calls for a combination of sophistica¬ted techniques and judgement - and making sure it has this mixture right is a key requirement for the NPD. One of the agency's principal jobs is to forecast future revenues from and costs for petroleum operations on the Norwegian continental shelf. Based on an annual collection and analysis of data, it predicts output, spending and environmental emissions/discharges - both in the short term and for many years to come. Look back a decade, to 1996, when the world was preoccupied with the re-election of Bill Clinton and Boris Yeltsin as presidents of the USA and Russia respectively. This was also the year when Britain's Prince Charles and Princess Diana divorced, the discovery of Mad Cow disease in the UK, the cloning of Dolly the sheep, and the Atlanta Olympics. Thirty-nine fields were in production on the NCS. Norway produced more than three million barrels of oil per day at a price of USD 20 per barrel, and sold more than 37 billion standard cubic metres of gas (scm) over the year. These were new records. The Yme field came on stream in the North Sea and remained in production until 2001. Development plans were approved for eight fields including Balder and Visund in the North Sea and Åsgard in the Norwegian Sea. Today, 10 years later, Norway is producing from 52 fields. The NPD expects oil production to be around 2.4 billion barrels per day and gas sales to total roughly 87 billion scm for the year. So what will the picture look like in 2016, a decade from now?
It is just as difficult to predict developments on the NCS as it is to foresee what the media will be reporting 10 years down the road. Questions include how many offshore fields will be producing at that time, whether the level of activity and production has declined, and whether major technological changes have occurred. Coming up with answers calls for a good grasp of history and appropriate routines for identifying trends and analysing plans from the operator companies. Every autumn, the latter are asked by the NPD to provide their forecasts for producing fields and plans for future development of proven discoveries. This reporting of petroleum-rela¬ted data for the revised national budget (RNB) forms the basis for the NPD's predictions of future petroleum production, investment, operating costs and emissions/discharges. These forecasts are submitted in turn to the Ministry of Petroleum and Energy and subsequently to the Ministry of Finance. Oil and gas operations mean a lot for Norway, accounting as they do for a third of government revenues and more than half the country's export revenues. Good predictions for the future level of activity are accordingly very significant for employment and economic policies, for example. Production volumes and development solutions will also be important for environmental emis¬sions/discharges and the possible future cost of emission trading or treatment measures.
The reporting process is accordingly extensive. This autumn, the NPD obtained information on more than 50 fields and an equal number of discoveries. Companies submitting reports have also increased in number as additional players become prequalified and receive interests in new production licences. Twenty operators reported to the NPD in 2006, and almost 100 types of data have to be provided, quality-controlled and analysed. The reporting system is built up from simple modules - the operators fill out standardised spreadsheets for transfer to the NPD's database. All types of data are reported at the same time to ensure consistency. Production profiles are analysed against reported costs, emissions/discharges and expected total recovery from the various projects. One of the most important tools used in processing the figures is the NPD's resource classification system, where all projects are classified by their maturity. Data uncertainty varies widely between sanctioned developments and possible projects which might be initiated in two, four or 10 years. Uncertainties relate not only to the volumes of oil and gas actually present but also to development concepts and the phasing-in of new fields over time. In addition to figures from the companies, the final forecast will also contain calculations for undiscovered resources. These are much more uncertain.
A good example of the challenges presented by the timing of developments is the forecast for crude production. Many Norwegian oil fields are in decline, with output contracting annually by up to 25 per cent. But work is under way on over 200 plans for improved recovery and to bring more than 50 discoveries on stream. Phasing-in these projects can have three possible outcomes, with one being that they start to produce in time to help overall Norwegian oil output to rise. If schemes are somewhat delayed, total oil production will stabilise at an annual average of roughly 2.4 million barrels per day. A third option is that projects are delayed or found to be insufficiently profitable, which will mean that overall Norwegian output con¬tinues to decline. The accuracy of the NPD's forecasts has varied over time. Cost overruns and delays affected its first attempts at prediction in the 1970s. Production forecasts were often pitched too low during the 1980s and early 1990s, when project after project for improved recovery was implemented on existing fields. That prompted the development of today's resource classification system. Major challenges currently relate to the availability of drilling units, steel, equipment and - not least - competent personnel to execute projects. In the longer term, oil prices and technology development will again become the issues which generate the greatest uncertainties for forecasters. The Tyrihans project in the Norwegian Sea illustrates how technological progress can help to create good and profitable developments. After this field was discovered in 1983, operator Statoil prepared a plan for development and operation (PDO) in 1997 but then withdrew it because of poor economics. A new PDO, involving extensive use of new technology and big improvements in costs and expected oil and gas output, was submitted the authorities in the summer of 2005. Production is due to begin in 2009. Demand for oil and gas, and thereby future prices, are the joker in long-term forecasts. During Norway's brief petroleum history, the cost of crude has ranged from less than USD 5 per barrel to its current level of roughly USD 60-70. Over this period, both the biggest field developments and major improved recovery projects have been sanctioned - regardless of short-term fluctuations in the oil price. Taking decisions on the basis of expected price trends over a long period is important for such major projects, which could have an operational life of 10-40 years. So prices applied in financial assessments do not match the current cost of oil. After the price slump in 1998, it took five years before the oil companies changed their decision price from well below USD 20 per barrel to USD 25-30. The NPD's resource accounts are to some extent price-independent - at least in a long-term perspective. The resources are in the ground. When developments are launched is a question of price and timing. Both technology and expertise in project execution must also be securely in place. Operations in deep water and the far north are examples of areas where technological advances have been made over the past 10-20 years.
The NPD forecasts include a prediction that annual Norwegian gas sales will rise over the next decade by more than 150 per cent from today's level, to 130 billion scm. Given current knowledge of the NCS, oil production will undoubtedly be lower by 2016 - and is likely to be about 50 per cent of its present level. And annual investment in the oil sector will probably decline by a third over the period. But these estimates remain uncertain. Although they represent the best present knowledge, history has demonstrated that even 10 years is a long time in the prediction business. The NPD has been responsible since 1993 for reporting from the petroleum sector. It devotes four-five work-years to this job, but 30-40 staff can be involved in this extensive task during the most hectic period every autumn. Information gathered by the NPD has many users in different areas. These data are utilised for in-house studies, and other players take frequent advantage of them. But facts collected by the agency are primarily applied by government ministries as the basis for White Papers, resource accounts and key publications about future developments.