You are Home   »   News   »   View Article

Getting more sophisticated with spares

Thursday, March 18, 2021

Oil and gas companies are getting much more sophisticated about how they manage spares, including maintaining optionality to buy a similar product from another supplier, rather than being constrained to only buying a spare with the part number in their systems, says Peter Hardy from Sphera.

In the boom times of the oil and gas industry, the approach to spare part management often boiled down to keeping as much spares as you could in stock, keeping simple databases of part numbers, and hoping you had the parts you needed when something broke, says Peter Hardy, director, of MRO (maintenance, repair, and operations) delivery with Sphera.

People had an incentive to hold as many spare parts as they could because the critical issue was not the money tied up in storing spares, but avoiding getting in a situation of having to explain why a certain part was not available. 'We would find whole warehouses full of stock which had never been touched.'

But now companies are getting much more demanding, aiming to reduce the parts they keep in stock without getting caught without something they urgently need, and also maintaining flexibility to switch supplier.

Sphera, based in Chicago, specialises in data and consulting services for heavy industries.

More nuanced strategy

To keep maximum optionality with minimum cost, you need a far more nuanced and integrated strategy with spares data, Mr Hardy says.

People are saying that if they are not using the parts in stock at least once or twice a year, they should not be holding them. Storage and capital costs can amount to about 20 per cent of the value of the materials per year.

But to be able to safely reduce stocks, they need predictions of which spare parts might be urgently needed, classifying spare parts for criticality, having skilled people who understand what parts are right for the task, and making it easier for engineers to find the parts they need.

Maintenance staff need to know about the function of the item. Inventory management departments need to know how they can store or transport it, and if there are any storage requirements.

Sometimes optimisation strategies solve one problem but create problems elsewhere, such as reducing spares in stock but being more reliant on fast deliveries, or having engineers spending more time trying to source the exact right part.

In order to better predict demand, companies are using techniques like Montecarlo simulation or actuarial science.

A further issue with spares is that much older equipment requires expertise to fix which is no longer available. So companies may choose to replace equipment rather than fix it, even if the spares are available.

Developing a spares strategy requires both data management and subject matter expertise, Mr Hardy says.

Master data

The underlying challenge is getting the 'master data' right, he says. That includes cleaning the data, ensuring it is accurate, and standardising it.

Wrong master data can lead to extra costs in multiple ways. For example, if you can't find the part you are looking for in the company stores, because the location shown in the computer system is not correct, you have to pay to buy stock which you already own, and there are delays waiting for deliveries.

If you have duplicate items, so more items in storage than you thought, that increases the cost further.

If you fit the wrong parts, there can be unexpected machinery failures, possible safety and liability issues.

Too much information or data floating in the system can also be a problem. An Accenture study found that the cost of each redundant piece of material in an ERP system can be about $30, in terms of making it harder for people to find what they are looking for, and then cleaning up the data.

High quality master data is also needed to negotiate contracts and determine the lowest cost suppliers.

Consistency of data is very important in helping people find the right parts - if you know how items are described, you know how to find them.

And not all spare parts are equal. A problem with a valve can stop operations, but a crack in the bezel of your monitor won't stop anyone doing anything. Data can convey the criticality of each component.

'The whole point around master data is around consistency, completeness and accuracy of information you are pulling in,' he says.

Functional specification

Different suppliers making equivalent spare parts are unlikely to agree to use the same part number. So in order to have choices about which supplier to use, a spare part needs to be described by its function, not just a manufacturer's part number. But there is not usually any standard way to describe items.

As an oil and gas real life example, you have five different pieces of equipment which all have what looks like the same valve. But the first valve is certified for a liquid up to a certain temperature - which makes it fine to use in plant A, but not on Plant B and C.

A company which supplied a complex piece of equipment, such as a pump, may try to steer customers to buying all spare parts from itself, although the spare parts themselves may be made by other companies, such as bearings. Sometimes, equipment suppliers will just cover the original manufacturer's part number with a paper label.

But there could be other issues involved - perhaps the machinery supplier does its own quality checks on the bearings and rejects 50 per cent of them. If you buy the bearings direct, you won't know.

Another reason to have a functional specification is in case a supplier goes bust. This is the only supplier who will be able to provide a part with its specific part number. Other suppliers may be able to provide an equivalent replacement part, but you need to know the part's specification to do that.

Mr Hardy used the example of work gloves to illustrate functional specifications further. These are available in a number of different materials, types, sizes, colours, construction methods. There are standards for factors like 'cut resistance'. They have lengths and sizes. Different specifications make gloves appropriate for different jobs. But there is no standard way to describe them.

Companies need to ensure that people have the right gloves for the right jobs, when doing machinery or maintenance work, but want to avoid spending more for gloves if they can. Oil companies purchase enormous quantities of gloves, so a few dollars saving on each pair mounts up.

Companies often want to have corporate purchase systems, where individuals in the company are able to buy what they need from a catalogue, restricted to purchasing items which have been approved by the company as suitable in both specification and price.

But commercial companies don't necessarily support the idea of a standard functional specification system, because they want to strongly encourage buyers to use their products. Some vendors in particular have a reputation for never providing any specification or information about their parts. 'The last thing any supplier any wants to put is any information on interchangeability. It has always been a fight,' he says.

Semantic web

Mr Hardy is very enthusiastic about the idea of a 'semantic web' for item functional specifications - which basically means items are tagged in a way that can be read by computers. It means a specific way to describe something - with nothing supplier-specific. A standard language for machines to read the web.

People can give items multiple names, but the computer only has one definition. 'I can call it one thing , you can give it a different name, we know how those things fit together,' he says.

There is an ISO standard for functional specifications, ISO 8000, which could be confusingly described as a 'standard for a standard'. Suppliers can provide a standard system for describing the functions of their products, which can be approved to see if it meets ISO 8000.

The ISO standards can be a tool for buyers, if they stipulate to suppliers as a condition of contract that data must be in ISO 8000. 'That's what will drive this,' he says.

If no standard exists, the supplier creates a metadata standard and submits it themselves into an approved ISO 8000 depository operated by another company.

Like all ISO standards, it doesn't tell you specifically how to do something, it says how it should be done. So it can describe how a metadata standard can be put together - then a standards organisation like PIDX, or anyone, can put it together.

About Sphera

Sphera specialises in data and consulting services for heavy industries, and acquired SparesFinder, a UK company focussed on improving quality of spares data, in 2018. The spares management work is within the 'operational risk' department of Sphera, reflecting that managing spares is perhaps best seen as a risk issue. It may be dangerous to issue someone with a 'permit to work' if they don't have the spares they need to do their work.

Sphera seeks to bring all elements together into one integrated environment, so its spares management tools are integrated with its permit to work management tools and barrier management tools, as well as master data management.

Associated Companies
» Sphera
comments powered by Disqus


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


Latest Edition Jan Feb 2021
Feb 2021

Download latest and back issues


Learn more about supporting Digital Energy Journal