Bearing up: finding parts for decades-old plant can be a nightmare. But Lee Hibbert talks to a company that makes hard-to-replace bearings for aerospace and nuclear uses.
The reason, of course, is that most of our nuclear power stations are now decades old and still retain much of their original equipment. These legacy systems are not just evident in the control room, they exist in many forms, both as hardware and software, and are all over the plant. There are no safety issues attached to this old age--in fact problems only occur when certain parts become obsolete and can no longer be supplied.
The military has the same problem. Some aircraft, such as the Harrier Jump Jet, have been in service for as long as 40 years. Although these aircraft have been upgraded time and time again, they still house certain components that were originally conceived many years ago. If the suppliers that made those products no longer exist, or no longer make that part, the issue of obsolescence rears its ugly head.
"Obsolescence is a problem in many industries, including military and aerospace, railways, oil and gas, nuclear and medical:' says Michael Trenchard, chief executive of the Components Obsolescence Group (COG), a pan-industry body set up to combat the problem. "It can potentially exist in any industry that maintains equipment for longer than the life of the components contained within.
"The best example I can think of is the American B52 aircraft. That was designed in the 1940s and, incredibly, will still be in service in 2040. Imagine how difficult managing obsolescence has been on that programme."
COG has more than 200 member companies, including giants such as BAE Systems, BP, AgustaWestland and Rolls-Royce. All of these companies have benefited from COG's resources which include education seminars, training and networking opportunities to enable them to manage the problem of obsolescence.
But, across wider industry, Trenchard says that too many firms prefer to stick their heads in the sand. "Unfortunately, I'd say the norm is still to be reactive rather than proactive when it comes to obsolescence," he says. "That doesn't make sense because, if you assume that problems will just go away, then you run the risk of being hit hard with idle machinery and equipment, and increased costs. The sheer rate of technological
change means that the numbers of drivers of obsolescence are increasing, and companies have to bear that in mind."
So what exactly can be done to manage obsolescence? Trenchard suggests the implementation of an obsolescence monitoring programme, the first stage of which is to understand your installed inventory. Only then can a database of components be created, and their status and a prediction of lifecycle be noted. Further assessment as to
whether they are single or multisource components can then be used to flag up potential problems.
COG also has a "who are they now?" database of companies that might have changed hands many times. That can prove useful in helping to track down suppliers and difficult-to-source components.
"Establishing an obsolescence monitoring tool isn't always a cheap or easy thing to do," admits Trenchard. "But, if you get it right, it can have a short payback time of between three to five years."
One facility that has already gone through this process is the Sellafield nuclear plant in Cumbria. Sellafield became operational in 1947, and equipment built then is still operational today. So obsolescence is taken extremely seriously and is managed carefully on a daily basis.
Graeme Rumney, senior technical services manager within the control systems department at Sellafield, is tasked with overseeing these activities. He says: "Obsolescence issues at Sellafield are many and varied. They tend to revolve around control systems engineering and, as a result, we are constantly replacing and revalidating the systems that we run.
"Particular items that need assessment include computer processing cards and memory cards, because industry moves forward very quickly with advances in these areas."
Sellafield has a site-wide asset management process and its obsolescence programme hooks into that. Rumney and his team regularly assess each system to ensure that a supplier exists, that they have enough spares, and that they have the required skills to support it. Each system is ranked and rated, enabling informed investment decisions to be made.
Rumney says that the system works well, but isn't foolproof. "Occasionally obsolescence issues around PC systems can creep up on us, but we have a good relationship with our original equipment suppliers and they always give us plenty of notification of when a product is about to become obsolete," he says. "That way we can consider what to do--either buy a large amount of spares or replace the system in question as a whole. I don't think a system has ever gone down here because of obsolescence and we haven't been able to recover it to a working state."
Rumney echoes COG's words when it comes to offering advice to other firms concerned by the risk of obsolescence. "Ignoring it could stop your facility in its tracks" he says. "Everything is so integrated in plants these days that an entire facility could easily go down and there could be a lot of lost income. Don't leave it too late and let your equipment fail, because your supplier might not be in a position to help you"
A mechanical component that can cause major obsolescence difficulties across plants like Sellafield and others is the humble bearing. Bearings are installed all over nuclear power plants, in equipment such as actuation systems, which position the control rods into the fuel bundle. In an emergency situation, the control rods are dropped into the fuel bundle to absorb the reactor heat, which dictates that the bearings must not, under any circumstances, fail. This means bearing suppliers must offer full manufacturing traceability, controlled lubrication and complete retention of records.
But it is quite possible that the original supplier of the bearing no longer exists or has relocated overseas and is difficult to track down. Or perhaps it is no longer willing to manufacture the equivalent bearing in small volumes of less than 1,000 units. In addition, there may be design features about the bearings that make it unattractive, or even impossible, for a supplier to provide them. For example, the bearing may require special materials, special lubrication, certification, special features such as tapered outside diameters, or may need to fit an Imperial chassis size.
For some firms, this sort of situation brings business opportunity. Plymouth-based bearings maker Barden has seen a growing demand for the manufacture of special configuration bearings in small batch sizes, anything between 10 and 500 units.
Sales manager Robin Kyte says: "The problem of obsolete bearings is predominant in the military aerospace and nuclear sectors. The problem usually arises when the mechanical system is very old, usually more than 25 years.
"In this scenario, many customers approach Barden because we are willing to manufacture direct replacement bearings in relatively small batch sizes and we have the manufacturing capability to produce special engineered bearings. Barden also have a strict non-obsolescence policy."
Often the bearing drawings on legacy products will define only the chassis size, with little or no information about the internal design, materials or tolerances. With such a basic specification, a wide range of bearings could meet the requirements, ranging from low-precision, commercially produced bearings to high-precision bearings with full traceability and inspection.
"Some of these bearings may be relatively inexpensive to buy, but they were never designed to operate in these types of environments or applications,' says Kyte. "Often, the buyer doesn't realise that the bearing they have just purchased is not a like-for-like replacement. Certification, full traceability and quality should be the key issues here. Barden provides full traceability, controlled lubrication and complete retention of records."
Obsolescence brings many different requirements to Barden's door. Requests for bearing sizes range from 5mm to 25Omm outside diameter. Most bearings manufactured for obsolescence requirements are either angular contact or deep-groove types, but the company can also now produce roller bearings.
Being able to produce bearings from special materials and coatings to suit the application or extreme environments has also proved useful. Balls can be made from silicon nitride, TiC-coated, zirconium dioxide or tungsten carbide, while cage materials can be specified in steel, bronze, phenolic, polyamide and polyimide. Lubricants include hydrocarbon, synthetic esters, silicone and perfluoroalkylether.
Barden is also capable of manufacturing bearings to a geometric tolerance of P2 or better, and envelope dimensions to P4 or better. Raceway roundness is better than 0.Sp, with raceway surface finish better than 1 micro inch Ra.
|Printer friendly Cite/link Email Feedback|
|Publication:||Professional Engineering Magazine|
|Date:||Oct 15, 2008|
|Previous Article:||Now you see it ... now you don't. Will we be able to make things invisible one day? Lee Hibbert reports on research on metamaterials.|
|Next Article:||A clean sweep: China's ever-growing appetite for fuel has brought an economic miracle and an environmental nightmare. Justin Cunningham reports as...|