To drill or not to drill?

Unconventional resource plays are game changers – for companies, for the industry, and for nations. Currently, about 75% of all oil and gas investment in North America is focused on unconventional resources. Because of unconventional plays, the US is heading toward energy self-sufficiency – something no one would have predicted 10 years ago. And while there is no denying these plays have been a driving force in the energy sector over the past several years, exploiting them has not come without pain. For many E&P companies drilling results have not lived up to expectations, and the financial impacts have been substantial.

“When unconventional drilling programs fail to deliver, the first reaction is often to tighten capital and operating costs. But in my opinion it’s often not a cost problem at all,” Boyd Russell, president and founder of Energy Navigator, said. “It’s a revenue problem. Companies that have not seen the results they expected likely never should have drilled in the first place.

“But the forecasting methods they had at their disposal yielded an incorrect production profile that led them to a wrong conclusion.”

Russell is an experienced reservoir engineer, and his Calgary-based company specializes in capital and reserve asset management software. He and fellow Energy Navigator engineer Randy Freeborn recognized early on that traditional forecasting methods were inaccurate for unconventional resources. In particular, they feel there are two key areas of evaluating unconventional reserves that require a fresh approach: decline curve analysis and type well analysis. They discussed their research into these topics in more detail in SPE paper 158867, published through the Society of Petroleum Engineers.

FIGURE 1. Several factors influence uncertainty in determining NPV. (Images courtesy of Energy Navigator)

“Unconventional reserves don’t respond the way traditional reserves do,” Russell said. “So why would we continue to use traditional forecasting and management methods that are simply not valid for unconventional reserves? When a company is planning to develop a large-scale play, management had better be sure of what it has before it makes a decision to start a drilling program that could negatively affect the company and its investors for years to come.”

How important is forecast accuracy?

In a 2010 white paper (SPE 135208), Haskett and Brown identified the main sources of uncertainty that companies face in determining the net present value (NPV) of a shale gas play. As shown in Figure 1, fully two-thirds of the uncertainty around the NPV of shale gas plays can be attributed to the reliability of the production profile, while variations in capex would have relatively little impact. This illustrates the critical role of the production forecast in decision-making and the ultimate success of the play.

“Whenever I’m presenting to industry audiences, I keep reminding them that this a revenue problem,” Russell said. “When companies consider that nearly 70% of their go/no-go decisions depend on the production profile, that’s huge. If capital costs represent less than 20% of the equation, even cutting costs in half won’t save them if their forecast is off. And we’ve routinely seen unconventional forecasts off by 50%.”

New decline curve and type well analysis

After several years and thousands of hours of research, Russell and Freeborn maintain that they have found a better way, and they have been presenting their findings to industry audiences around the world. Their new approach to decline curve analysis and type well creation yields consistently reliable forecasts, accurate to within a few percentage points.

For decline analysis the pair’s recommended approach, called a five-year equation, uses traditional Arps equations for the first five years and transitions to boundary-dominated flow thereafter. For type well creation Russell and Freeborn recommend moving away from the current industry standard practice of using only historical data for analysis. Instead, their research shows that by using a combination of well history and forecast data for each well, the resulting type well is not biased by the historical development of the field and results in a much more accurate approach.

“We’ve eliminated the uncertainty inherent in traditional methods,” Russell said. “This method also eliminates the potential for sequence bias – the tendency to drill the best wells first and then base forecasts on those wells, resulting in a grossly overoptimistic production forecast.”

To prove the validity of their new methodology, the pair conducted an extensive blind back-casting project involving hundreds of wells throughout all the major basins. The back-casting results matched actual documented well results almost perfectly.

FIGURE 2. The results of blind back-casting studies showed a good match with real production data.

Auto-forecasting

But simply being able to find a more accurate approach only partially solved the issue of forecasting unconventional plays. Many factors influence each well in an unconventional play. Evaluators cannot simply rely on nearby wells with matching geological properties to establish a reliable analog. And the sheer number of wells in a play makes it impractical to forecast them all manually.

So while forecast accuracy is likely the most important success factor in unconventional resource plays, the second most important factor is the ability to auto-forecast quickly while still maintaining accuracy. Freeborn and Russell developed an auto-forecasting algorithm that is built into Val Nav, Energy Navigator’s integrated forecasting, economics, and reserves management tool. This auto-forecasting capability allows an engineer to forecast thousands of wells in a day with pinpoint accuracy – a task that could otherwise take weeks or months.

“We made it do what any experienced engineer would naturally do,” Russell said. “Val Nav’s auto-forecasting tool recognizes multiple trends within historical data, deals with inclining trends, identifies trends among bad data points, distinguishes between normal well noise and bad data, and handles gaps in production. When companies are basing multibillion dollar drilling decisions on their production forecasts, they want accuracy.” By completely rethinking and challenging the conventional evaluation approaches, Russell and Freeborn maintain they have found a better way to forecast and manage unconventional reserves. Armed with timely, accurate information, engineers can focus more of their attention and expertise on engineering and less on cost accounting while still achieving higher return on investment on unconventional resource plays. It is hoped that the new forecast methodology will have a positive impact on the level of success that companies can achieve in unconventional plays by helping them make better decisions about whether to drill or not to drill. That, in turn, will help investors and the financial community make investment decisions with much greater clarity and confidence.