Palantir Solutions – modelling your company in one go
Feature Articles, May  29  2009 (Digital Energy Journal)

- UK company Palantir Solutions has launched a new version of its economic modelling and financial forecasting software running on 64 bit which can handle data sets 3,000 times larger – so you can see what the impact will be of reducing production on a single well for a month, on the company’s annual tax calculation.

UK oil and gas financial modelling and planning company Palantir Solutions has beefed up its software to run on 64 bits, which means it can process data sets of up to 8 terabytes in one go. The software is being rolled out by its offices in London, Aberdeen, Calgary, Houston, Singapore and Bangalore.

This means that the computer system can process all of the variables involved in an enormous oil and gas financial planning model – without saving part of the calculation to disk along the way.

For individual wells, there can often be over 1500 variables included in financial planning – including how you expect production to change over time, what work you expect to do on it, who will own it, what royalties you have to pay, drilling costs, operating costs, labour costs all with perhaps P10, P50 and P90 variables. For many of these, you might have a new variable for each month over a period as long as 60 years.

Meanwhile for the whole company, you have variables for the oil price, and you might have to borrow money, at varying interest rates. You might have different tax systems which kick in at different profit levels.

Bryan Dyck of Palantir Solutions

Trying to forecast the future and work out how certain changes might impact the overall profitability can become an enormous calculation, too big in fact for a computer to store in its memory at once, says Bryan Dyck of Palantir Solutions.

If the calculation gets too large, then the computer system has to split it up into smaller chunks, which means it takes much longer to get the right result.

So Palantir has upgraded its financial software to run on 64 bit Windows, running on high performance computing, so it can handle data sets up to 3,000 times larger. Theoretically, the computer can handle datasets of 8 terabytes, compared to a previous limit of 2 to 3 gigabytes under 32 bit Windows.

This means that a company can remodel its entire business plan with each small change (eg a decrease in rig rates or a change in the oil price).

The 64 bit capability was released on May 27th as part of its version 3.5 software upgrade. Companies already using the new software include Talisman, Dana Petroleum and E.ON Ruhr Gas

TAQA.

Finer granularity

In the current business environment, many customers are keen to get an understanding of what impact small changes might make on their overall business, for example finding cost efficiencies, or working out the best place to build a new hub, or what is the cheapest way to increase the reserves on their books, or which assets are the best to sell.

“Having the data to a much finer granularity will give much greater insight on which assets are bringing them down,” he says.

“Before this change – there was always a trade-off between the granularity and the accuracy of the data. You could have accurate data, but process less data. You couldn’t have both.”

So, for example, companies take loans and pay tax and see the financials on a company-wide basis – but they don’t connect these decisions easily with decisions about individual wells.

“With our system we know how much of the corporate loan could be attributed to each individual project.”

The financial models often have to be auditable, so someone can see later exactly how a decision was made. “There’s always a lot of data you have to process,” he says.

“It gives us that ability to really push the limit – to address the data and the business workflows in a way we couldn’t previously.”

The software can be used to see what your cash flow and profitability will be over the next 40-50 years based on your current assets and how you expect your asset portfolio to change, looking at data for each month.

Cloud computing

As the software calculation gets bigger, it is likely to be moving off desktop computers into corporate high performance computer rooms, or remote “cloud” computing systems. “The next thing that we’re focussing on – is pushing this into a cloud environment,” he says.

Palantir could host the cloud computing network itself, or rent computing space from a larger company; or oil and gas companies might develop their own cloud computing networks.

“Large calculations are better on a strong centralised environment,” he says.

Palantir

<<BACK