You are Home   »   News   »   View Article

Validere - understanding hydrocarbon quality using analytics

Thursday, December 17, 2020

Validere Technologies, based in Houston, Calgary and Toronto, provides an analytics based service to advise customers about the quality of their flowing hydrocarbons.

From the well to the customer, hydrocarbon flows come in a wide range of specifications and quality levels, and it is very useful for people involved to know more about the quality.

But gathering together available data and making inferences about it is a very complex technical and analytical challenge. This is what Validere Technologies, based in Houston, Calgary and Toronto, is tackling.

The company was founded in 2014, so just about qualifies as a start-up. According to it has $25.7m total funding in 6 rounds. Major investors include Wing VC, Greylock, and Sallyport, which also backed energy technology company RigUp.

The platform is priced by monthly fee.

Validere builds models of available data, validates it, and then 'pushes' out insights to the client.

Validere also has an online platform, 'Validere Edge', which connects together buyers and sellers of commodities, based on a need for products of a certain quality.

Data challenges for producers

Oil producers need oil to meet a certain specification in order to be transferred onto a buyer. If the spec is outside a certain range, the oil cannot be sold under this contract.

There are various ways to take and measure samples, including desktop devices and sending samples to a laboratory.

Fitting an instrument 'inline', so it can automatically take samples out of the pipeline, can cost $500k, says Mark Le Dain, vice president strategy at Validere.

Sometimes companies have instruments which are not calibrated correctly or which go out of calibration quickly. There can be errors in data gathering and recording.

And oil shale wells in particular have high decline rates, and the specification of the oil tied into a system can quickly change as the reservoir depletes and is replaced by new production.

Supply chain data challenges

Companies involved in the oil and gas supply chain, and midstream operations, can also benefit from product quality data in many ways. They can use the data to optimise their own processes and make sure there are no problems when hydrocarbons change ownership.

When oil changes ownership, the important data points are the price, volume and the specification, or quality. But data about the price and volume is far easier to gather than data about quality, Mr Le Dain says.

But companies can pay big penalties or costs if they try to supply products which are outside their customers' specification.

For example, if they load out of specification crude onto a truck, it can get refused at a terminal, so they have to pay for it to be delivered somewhere else. There have been stories about entire pipelines getting contaminated after some crude of the wrong specification was pumped into them.

Often volumes are mingled together, making it even harder to track what is being bought and sold.

The company's service is not limited to helping customers analyse only their own data, as it helps them interact with the supply chain most efficiently. . So it can track product quality as it is handed from one owner to another, provided companies give permission for their data to be used in this way.

This way, refineries and intermediaries purchasing crude can use the service to get a better understanding of what they are buying.

Consider that a maritime shipping company might want to avoid using a fuel supplier it does not know, because of risks that it could be provided with substandard fuel with compatibility risks.

With better information about the provenance of this fuel, a company could be able to buy fuel from a previously unknown, but lower cost supplier with more confidence, Mr Le Dain says.

Other benefits from having better data include identifying leaks or theft in hydrocarbon flows, optimizing margins, getting a better understanding of markets, and resolving custody disputes.

How to analyse data

The starting point of data analysis is to gather together all available data - including any 'inline' devices, sensors and sample test results.

There has always been a large amount of data available about fuel quality, but is often 'scattered around, incorrect, or not actually used in economic decision making,' Mr Le Dain says.

The available data is used to put together a physical model about the liquid properties. Many data errors or inaccuracies can be identified from the process of putting this model together. When you find that data doesn't fit the model, something must be wrong somewhere. For example, it can show that fluids which were moved from tank A to tank B can't possibly have the properties that the sample says that they do.

This can lead to additional testing, or an understanding that there is a flaw somewhere.

Validere employs experts in both fuels and data science to put together these models. Co-founder Ian Burgess has a Ph.D. in Applied Physics from Harvard.

Where there is data which would be desirable but is not available, the modellers use engineering and physics models to try to work out what this data would be if there was a sensor to record it. This process is sometimes called developing virtual sensors.

For example, a physics / engineering model may describe a relationship between viscosity and density, or what the vapour pressure would be.

Sometimes these virtual sensors have actually given better data than inline instruments, because inline instruments need regular calibration.

Validere also develops its own engineering models, based on cases where it does have the data from sensors. So (for example) it can directly measure both viscosity and density for many different oil samples and build a model about how they relate in different conditions.

Ultimately, as a result of the data gathering and modelling, it has much better data to provide to the client, for use making supply chain decisions.

The challenge of building systems to better gather and manage fuel data can be much more about implementation than designing clever solutions, Mr Le Dain says.

Associated Companies
» Validere
comments powered by Disqus


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


Latest Edition Dec 2020
Nov 2020

Download latest and back issues


Learn more about supporting Digital Energy Journal