Digital Energy Journal- Tessella's 'analytics partnership'

Join | Existing Member Login

You are Home » News » View Article

Tessella's 'analytics partnership'

Friday, August 29, 2014

If you might like some data analytics done, Tessella offers long term 'partnership' arrangements - so you can have an analytics expert available on standby

International data analytics company Tessella is offering oil and gas companies 'partnership agreements', where they can have advanced analytics specialists always available to assist with solutions or just offer advice, under an on-going retainer agreement.

The aim is to make it easier to access a consultant on any specific project, by lowering the barrier to making that choice.

'If we have a partnership agreement in place, each client has the continuity of a dedicated contact within Tessella who really understands their business needs,' says Nick Clarke, Head of Analytics at Tessella.

'We don't need to start a new conversation or draw up a new contract every time for every problem. We work together to identify and provide the exact capability needed in each case.

At the moment, it can be difficult and timeconsuming for oil companies to assess whether they need to engage a consultant or can deliver an analytics project themselves.

The partnership approach stops client and supplier having to do that dance around the detailed specification of a problem and associated deliverable before anyone gets going, which can be a big frustration,' he says.

'The potential value from analytics can easily be lost if it cannot be delivered with enough speed, and the last thing you need is to be stuck in contracts while your competition marches forward.'

Tessella now employs 250 consultants with a wide range of domain expertise, including condition monitoring, image and acoustic analysis, advanced statistics and reservoir modelling. The partnership agreements give clients an easy and flexible route to accessing that full breadth of capability, when they need it.

Empirical vs engineering models 'Analytics is at its most effective when you can use a mixture of observed data together
with engineering models to figure out what is happening now, and predict the most likely future,' Nick Clarke says.

'It's the combination of the two that gives you real predictive and interpretative power,' he continues. 'To come up with a reliable interpretation of what's really going on, you need an underlying model you can track measurements against. It's how you get from correlations in the data that might mean something, to hypothesising causal relationships you can rigorously test for.'

Achieving this with sufficient accuracy in remote or hostile environments can be a real challenge. As an example, UK nuclear waste processing plant Sellafield had a critical piece of machinery which required a metal bar to bounce freely inside a highly radioactive rotating drum.

Tessella consultants created an idealised model of the acoustic signature (sounds) that would be expected from a bar in good condition.

They then fed the actual sounds (which were recorded in a very noisy environment) into an analytics engine that compared the evolution of the recorded sound with that expected by the model. Taking into account all of the combined uncertainties inherent in the idealised model, the background noise and the microphone (this was the clever bit), the analytics engine calculates the probability that the measured sounds indicate the bar is beginning to deform.

A wide variety of skills was required to develop a viable solution, and providing fast and easy access to them is what the analytics partnership is all about.

Associated Companies
» Tessella

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.