One thing he has reacted to during his long experience with oils and liquid products in the power industry is that there is a lack of knowledge about the most central issues one needs to keep track of.
– When I became an independent consultant, after many years at ASEA/ABB and Vattenfall, and was hired as a lecturer for courses around the world, I was struck by how very few people knew about, for example, the Arrhenius equation that Svante Arrhenius, Sweden’s first Nobel Prize winner, published in 1886. This surprised me, since calculations based on it are required to assess the remaining service life of a system and an oil, and every chemist studying at university level has to do at least one lab on the Arrhenius relationship, he explains. How can it be that these relationships are “unknown”?
Degradation rate determines service life
The Arrhenius equation shows the relationship between temperature and the speed of different reactions, something that is absolutely decisive for the service life of an oil. The oil may already be 10–100 million years old when it is pumped out of the ground, which immediately proves that contact with oxygen in the air is the most fundamental cause of oil consumption/degradation, i.e. oxidation processes.
– For example, an oil oxidises twice as fast with a temperature increase of 8 degrees Celsius, but for the sake of simplicity we usually say 10 degrees. In a transformer, where cellulose paper is used as dielectric insulation on the windings, it is more important to keep track of the degradation of the cellulose because there the degradation rate doubles for every increase of 6 degrees.
When Lars Arvidsson began working with insulating oils and lubricants at ASEA, Swedish Standard required that only naphthenic oils be used in transformers, because at that time one might have to cold-start a transformer at -50 degrees and then the oil had to remain fluid. The choice of naphthenic base oil was actually a coincidence. Vattenfall set the requirement for the insulating oils, and the leading manufacturers sent samples to Vattenfall’s materials laboratory in Västerås, where they were all rejected because the lowest pour temperature was only -20.
Nynäs then suggested testing one of their refinery fractions, which at the time had no market, and that oil proved perfect because the naphthenic product remained fluid down to -55 without additives. Since then, Nynäs has, through skilful marketing, created the image that a transformer must have naphthenic oil. At various conferences he has heard, for example, that when Westinghouse built his first transformer at the end of the 19th century, he chose naphthenic oil for its good properties. But at that time no one knew what naphthenic oil was. At another conference he heard that PCB forms in transformers.
– Suppliers of naphthenic oil claimed that only such oils worked, but in reality the only requirement for a transformer oil is that it must be electrically insulating. That means even ordinary kerosene or diesel would work. However, kerosene and diesel are very volatile and therefore highly flammable, making them unsuitable in practice, but in catastrophic situations you can pour such liquids into a transformer without anything happening.
A by-product from Nynäs created a global trend
The fact that naphthenic oils are now recommended around the world is thus due to a development in Sweden during the 1950s.
When hydropower, led by Vattenfall, had been fully expanded, Nynäs lost its large customer group and began targeting other markets, especially the British market because all former colonies used BS, British Standard. Soon naphthenic oils became standard all over the world as well.
– But countries that do not have low winter temperatures do not actually need naphthenic oil; they can use paraffinic oils instead. Yet manufacturers of naphthenic oils have continued to market naphthenic oils as the only alternative. A repeated lie eventually becomes an unquestioned “truth”.
Far too little knowledge about oils among users
Lars Arvidsson uses the historical account of why naphthenic oils were long required in transformers as an example of how users of oils in different contexts have far too little knowledge of oil properties and of what performance should actually be demanded. The other day they analysed the oil from a hydropower plant and found that it had actually been intended for a gas turbine. Apparently, the machine owner had called the oil company and asked for turbine oil, and since the oil company had not developed any specific product for hydropower, the result was an oil optimised for temperatures above 100 degrees Celsius. A hydropower turbine can in fact run on a good base oil with no additives other than an antioxidant.
– Many companies rely far too heavily on suppliers. One example of this was when the first generation of ester oils appeared.
At the time he worked at Vattenfall and was responsible for its test laboratory, and they chose to conduct their own tests before they would even consider using the new “environmental” oils.
– When we tested the first generation of ester oils in the 1990s, we saw that they had fantastic lubricating properties, but they absolutely did not have stable properties over time because they began to break down immediately in the standard tests we used back then. In a transformer, for example, the viscosity must remain constant for many years, which esters definitely did not. In one test where we required that the oil should not change viscosity over 1000 hours, the oil had turned into gel after 25 hours.
Vattenfall chose to develop a cleaner oil itself than it had previously used in order to meet the environmental requirements.
– We chose to use white oil, a product already on the market that had been developed for the food industry and was therefore approved for food contact. That gave us a much cleaner oil with the same properties as ordinary mineral oil but without the disadvantages that the esters offered.
Around that time Vattenfall decided to shut down the laboratory where he had worked. He then chose to start his own company with the same function, what has now become VPdiagnose.
– We got many assignments abroad; among other things we began a collaboration with Saudi Arabia, where we performed oil analyses for them. I have also given hundreds of courses around the world, primarily on transformer oils but also on hydraulic fluids. Our strength is our level of knowledge combined with an independent business. We do not collaborate with any manufacturer, so people can really trust that we do not issue statements in order to sell any other products. Our only product is unbiased knowledge.
Experienced problems with the hydraulics industry
Lars Arvidsson chose relatively early on to focus on transformer oils because he felt there was a greater understanding there of what an oil actually is.
– My experience of the hydraulics industry is that many people there think oil behaves like water when it flows around. The big problem is that there is a lack of knowledge of fluid dynamics in that industry. In the 1990s, opinionated people dominated. They presented views that lacked a scientific basis but were still very influential in the standards committees, so one felt very “alone”, since the chemists who were there were employed by the oil companies and naturally wanted to sell an oil with the price-raising label “environmentally adapted”, which also wore out much faster and could not be recycled. Mineral oil, by contrast, could be re-refined any number of times, but the oil companies did not want the requirement to handle waste oils.
An important insight he wants to convey is that a hydraulic system should be designed to suit the oil, instead of first building the system and then pouring in an oil.
– This is mainly about ensuring that the system allows the oil to de-aerate. Mobile systems, for example, should be as small as possible while still allowing the oil to release air. One question I ask myself is how often a hydraulic technician calculates the flow velocity in a system.
At VPdiagnose they had a customer who needed help with a hydraulic system, and among other things the flow velocity in one of the hydraulic pipes could be calculated at 90 km/h. No wonder problems arise.
– That gives the oil a very high momentum, which causes problems for example in on/off valves. When the valve closes, the oil continues by inertia and is then sucked back, causing implosion against the valve, which shortens the life of the valve but also creates an adiabatic compression of the gases that have come out of the oil during the motion. The temperature there can reach above 500 degrees, i.e. hotter than the operating temperature in a refinery cracker, where the purpose is precisely to break oil molecules into polymerisable components. Knowledge that it is easier to release gases from oil than to dissolve them back in is poor.
Gas analysis is a useful tool
Understanding how the oil flows through the system is important in order to understand how air bubbles behave so that the best possible deaeration can be created.
– Oil is more viscous than water, which means it takes longer to release gases and air. You have to create a flow pattern that allows the air bubbles to move “northward” rather than swirling downward into the oil tank, which happens when the flow becomes turbulent rather than laminar. I have seen oil tanks with standing waves.
He notes that with a good gas analysis, it is possible to obtain information about how well a system has been designed.
– We have been involved in a number of cases where hydraulic systems had been delivered and, by means of gas analysis, it was established that either the system had been wrongly designed or the wrong oil had been used.
– I still see gas analysis as a fantastic way of quality-assessing a design.
