Managers have long complained about the cost of data and information management, of which companies see no measurable return. There may be ways of measuring the actual value.

Why manage data in the first place? If money is to be spent managing data, there must be a reason for it.
In the pharmaceutical industry, there is a regulatory need to manage data. Testing information is vital if a drug is to be released into the marketplace. In the exploration and production industry, the management of information also is regulated, but regulators rarely bother to follow up. For example, the UK government legally can ask any licensed company operating on the UK Continental Shelf to provide any data in any format at any time. The fact that the government generally chooses not to ask, is willing to wait when it does and historically accepts anything it is given has led to complacency.
Another reason for managing data is the creation of value, either for a company or a country. Exploration and development of hydrocarbon resources is inherently reliant on data, information and knowledge. In addition, certain companies are in the business of gathering data - whether from the field or by piecing together other people's data - which they then sell. These include service companies like Schlumberger, Halliburton and Baker Hughes as well as data vendors like IHS, Robertson and A2D.
Traditional value measures
The time lost by technical staff looking for data is the most commonly quoted measure of the value of data management. The most quoted paper on this dates back to November 1991, when Lee Lawyer of Chevron published a pie chart showing how the average geoscientist spends the work day. The figure then was 60% of the time looking for data and 18% spent on useful work. Has this changed significantly? In a presentation at a 2001 Schlumberger Forum, Smalley and Espeland came up with a figure of 44% of a BP geoscientist's time spent finding, accessing, quality checking and manipulating data and 50% spent on the value-adding phase of interpretation and analysis. The remaining 6% of time was devoted to documenting and archiving (Figure 1).
Other studies show the time and cost of finding data vary considerably, from 75% of time (Figure 2) to 14% to 15% of cost (Figure 3). The problem with many of these studies is that they measure different things and hence, are not comparable. However they point to the same end - more effective data management results in more effective use of time by geoscientists and engineers. But there is a limit - data management will never be free, and geoscientists and engineers always will spend some time looking for data.
Other measures
Another way of looking at the value of data management is to look at the cost of mistakes, lost production and loss of investment, as illustrated in three case studies.
Case study 1: Management buy-in.
In this case the company left data management to individuals and had no coherent policies in place. The management team began a cleanup and installation of data management systems. Once it began the process, it realized the problem was larger than anticipated, and it had no idea what to do upon completion. The company asked Paras Consulting to come up with an economic case for continuing investment.
The company's internal studies showed that time spent finding, loading and conditioning data took about 75% of the geoscientists' and engineers' time. Typical problems included:
• no one in the organization held responsibility for data management;
• two databases, one in the exploration department and one in the production department, showed the same wells deviated in different directions;
• wells had been twinned because the original well had been loaded in the wrong place;
• a deviated field well had been plotted in the middle of open acreage, several kilometers away from the field (the geologist just moved the well to the same location as a vertical well); and
• production data were reported differently in different places. The most extreme examples showed that the total daily production reported to senior management was significantly greater than the amount of oil arriving in the stock tanks.
This project had a life of 1 year, and it was difficult to see how the benefits would accrue beyond that time. The general opinion of the company's staff was that once the project was finished, everything would revert to "normal," and the money spent would be wasted.
The consultants built several models, based on different assumptions, to determine how the project could generate a return on capital investment. The base case was taken to be nothing would change, and the situation would worsen after adding hardware and software bills to the operating costs. The addition of a data manager would not help much, as a single person would be unlikely to succeed in changing things without a good organization in place.
The most likely chance of success was to finish the cleanup of data and put a data management team in place. Calculations showed that the return on investment was robust no matter which scenario was used. This approach won the approval of senior management, but failed when it reached top management, which measured performance in barrels of oil produced rather than in profits made.
So this company is continuing to throw about US $10 million of shareholders' money down the drain each year (a conservative estimate). The lesson learned is to make sure you know what drives top management.
Case study 2: Lost production. In this case an onshore US-based production company was experiencing difficulties with keeping track of its wells, especially when they lost production. The issue centered on knowing when a well or wells went down and what to do about it.
Various well engineers had information about each well in various formats in various places. If the engineer was away for any reason, it could take up to a month to find the information. Sometimes it took that long even when the engineer was present. When the company received information about the failed well, it questioned what might have happened in other nearby wells. Again, a search for relevant information could be time-consuming. During this time, production from the original well was lost. For a well producing 100 b/d, the cash flow loss during a month (30 days) with oil prices about $20/bbl would be $60,000. For many companies this is significant, especially as more wells fail in mature fields because of aging equipment.
In this case the company decided against continuous monitoring of its wells; instead it went for a Web-enabled system to index relevant well information. This resulted in an estimated 50% savings in downtime or a $30,000 increase in cash flow per well. Management did not need any further convincing to implement a solution of this nature.
Case study 3: Banco de Informacion Petrolera, Colombia. Colombia is well known for all the wrong reasons, which tends to discourage investment. To counter this, national oil company Ecopetrol developed new fiscal and operating conditions ("Colombia finds formula for outside investments," Hart's E&P, June 2002, p. 92). Discussions with operators showed they had learned to deal with security issues and that the fiscal regime was not problematic; it was access to reliable, high-quality information that caused some of the biggest headaches. Typically it could take up to 1 year for a company to sort out the information provided by Ecopetrol at the beginning of a contract.
Ecopetrol started off by cleaning up a significant proportion of the data, spending $30 million to $40 million in the process. It contracted Schlumberger to put a data management solution in place that could deliver data to prospective foreign investors with a minimum of fuss so visits were not essential. The solution has been up and running for a while, and the net result, combined with an increased marketing effort, is an increase in the number of licenses signed. According to Ecopetrol, "The more eyes that see its information, the better chance it will have to contract properties." The use of a data repository and data cleanup has put Ecopetrol in a more competitive position, and the change has helped overcome some of the negative sentiment toward investment in Colombia. The fiscal value of the project has not yet been measured, and it may be some years before this can be done sensibly.
Making the business case
These case studies prove a value can be placed on data management. The level of detail required will depend on management's view. The most important lesson learned is that it is vital to find out what drives management's decisions and how data management can support them reaching their targets. Each business case is different - there is no "one size fits all" in data management.

Comments