The analysis of gases from petroleum products has been performed for decades using gas chromatography. However, this technique was not applied specifically to transformer mineral oil until the late 1960s/early 1970s and is now commonly called dissolved gas-in-oil analysis (DGA). Some of the early developers of the technique were Dr. James Morgan of Morgan Schaffer Systems, Canada, and researchers J.E. Dind, R. Daust and J. Regis from the Canadian utility Hydro-Quebec.1
Because the technique was so successful and provided a wealth of diagnostic information to detect incipient faults, other laboratories such as Doble Engineering in Massachusetts began utilizing the technique shortly thereafter. DGA has now become a standard in the utility industry throughout the world and is considered to be the most important oil test for insulating liquids in electrical apparatus.
More importantly, an oil sample can be taken at anytime from most equipment without having to take it out of service, allowing a “window” inside the electrical apparatus that helps with diagnosing and trouble-shooting potential problems.
Insulating fluid analysis is performed using an array of physical, chemical and electrical tests. Physical tests include analysis for interfacial tension, pour point, relative density, viscosity, color and others. Tests such as water content, neutralization number, oxidation inhibitor and polychlorinated biphenyls (PCBs) are considered chemical tests. Electrical tests consist of analyses for dielectric breakdown voltage and power factor.
There are many other tests that can be performed but these are the main ones executed on a routine basis. Most tests performed in North America are based on ASTM test methods. In Europe and other parts of the world, similar test methods come under the auspices of the International Electrotechnical Committee TC10.
Because oil and solid insulation degrade with service in electric apparatus, periodic sampling and testing are necessary to ensure that deterioration is detected before it becomes excessive. The rate at which the insulating materials degrade depends on several factors, such as the type of oil preservation system (amount of oxygen present), operating temperature, water content of the insulation, and the amounts and types of contaminants.
The frequency at which testing is conducted varies with the test, importance of the equipment, whether an incipient-fault condition is known to be present or a problem exists, and when a family of transformers has been identified as having a history of problems.
The tests mentioned above provide information on the oil quality itself but fail to provide in-depth diagnostic information on the operating condition or health of the electric apparatus. In addition, because most transformers in the United States are considered to be sealed systems (equipped with preservation systems that retard the ingress of moisture and oxygen), oil degradation can occur quite slowly over many years.
Many of the transformers in the United States that have been in-service for more than 30 years still contain oil that is in good condition. There is one test however, also performed on the insulating oil of the transformer, which provides a means for diagnosing the operating condition of the electrical apparatus, namely DGA.
Dissolved gas-in-oil analysis, performed in accordance with ASTM D3612 or IEC 60567, is by far the most frequently requested diagnostic test and the single most important test performed on transformer oil. As the insulating materials of an electrical apparatus, such as a transformer, break down from excessive thermal or electrical stress, gaseous byproducts form. The byproducts are characteristic of the type of incipient-fault condition, the materials involved and the severity of the condition.
Indeed, it is the ability to detect such a variety of problems that makes this test such a powerful tool for detecting incipient-fault conditions and for root-cause investigations after failures have occurred. Dissolved gases are detectable in low concentrations (ppm level), which usually permit early intervention before failure of the electrical apparatus occurs, and allow for planned maintenance.
The DGA technique involves extracting or stripping the gases from the oil and injecting them into a gas chromatograph (GC). Detection of gas concentrations usually involves the use of a flame ionization detector (FID) and a thermal conductivity detector (TCD). Most systems also employ a methanizer, which converts any carbon monoxide and carbon dioxide present into methane so that it can be burned and detected on the FID, a very sensitive sensor.
Removing the gas from the oil is one of the more difficult and critical portions of the procedure. The original method, now ASTM D3612A, required that the oil be subjected to a high vacuum in an elaborate glass-sealed system to remove most of the gas from the oil. The gas was then collected and measured in a graduated tube by breaking the vacuum with a mercury piston.
The gas was removed from the graduated column through a septum with a gas-tight syringe and immediately injected into a GC. In the present modern day laboratory, however, mercury is not a favorite material of chemists. For this reason, two additional extraction techniques have been developed to eliminate mercury.
ASTM D3612B is called the direct injection technique. In this method, the stripping of gases from the oil and the gas analysis takes place inside the GC. Originally developed in the mid-1980s for this application, the process involves injecting the oil into a sample loop in the GC. When the GC run is initiated, the sample loop transfers the oil through a series of valves into a stripper column.
The stripper column is composed of metal spheres in one end in which the oil overlays the surface of the spheres to increase the surface area. Carrier gas is passed over the spheres and extracts dissolved gases from the oil, which then pass through a series of columns and on through the detectors. The oil is back-flushed and purged from the system before the next sample is introduced.
The newest method, ASTM D3612C, was approved about a year ago and is called the headspace method. Headspace technology was used for DGA analysis for almost a decade. However, it was never developed into a robust, reliable standard method until several years ago when Jocelyn Jalbert of Hydro-Quebec developed a headspace method using a Hewlett Packard (now Agilent Technologies) instrument.2
The technique involves injecting an exact volume of oil into a purged and pressurized headspace vial. The gas in the oil is then allowed to develop an equilibrium with the vial headspace under shaking and heating conditions. After a predetermined sample extraction time, the autosampler removes a portion of the gas from the vial headspace and injects it into the GC. The advantage of this method is that it can be automated and reduces the risk of operator error from excessive handling of the sample during preparation and injection.
Of course, each method has its advantages and disadvantages. Method A (ASTM D3612A) is by far the longest-standing technique and is still widely used today and offers accurate, reliable sample preparation if followed rigorously. However, the alternative methods, which are more easily automated, are gaining acceptance as they are shown to be reliable. None of the extraction techniques completely removes all the gases from the oil.
This is due to the solubility coefficient of each gas, which must be accounted for in the final concentration determination. Laboratories must also work with commercial suppliers to develop gas and gas-in-oil standards or they must prepare standards themselves, because these are not currently available from national standardization bodies such as the NIST.
Repeatability and accuracy are also of the utmost importance as small changes, even several ppm in some cases, can mean the difference between an active incipient fault condition that requires immediate attention or one that is stable and requires no attention.
Just like with industrial oil analysis, good sampling practice is important for obtaining accurate DGA data. Key gases such as hydrogen and carbon monoxide could easily be lost from a sample because of their low solubilities in oil. In order to minimize the loss of gases, ASTM D3613 requires samples to be taken using gas-tight glass syringes or metal bulbs; these are available from several suppliers or from commercial transformer oil analysis labs.
Typical gases generated from mineral oil/ cellulose (paper and pressboard) insulated transformers include:
Carbon Monoxide, CO
Carbon Dioxide, CO2
Additionally, oxygen and nitrogen are always present, their concentrations vary with the type of preservation system used on the transformer. Also, gases such as propane, butane, butene and others can be formed as well, but their use for diagnostic purposes is not widespread. The concentration of the different gases provides information about the type of incipient-fault condition present as well as the severity. For example, four broad categories of fault conditions have been described and characterized in Table 13.
Electrical discharges or inadequate cooling of the paper insulation cause it to overheat, generating carbon oxide gases. Examination of the relative composition or ratios of gases present can provide further refinement of the diagnosis. This typically involves using either Rogers ratio or Dornenberg ratios.
The severity of an incipient-fault condition is ascertained by the total amount of combustible gases present (CO, H2, C2H2, C2H4, C2H6, CH4) and their rate of generation. Generally, transformers will retain a large portion of the gases generated and therefore produce a cumulative history of the insulating materials’ degradation. This is an important tool for detecting and trending incipient problems.
However, it also means that care is needed in interpreting values for a first-time analysis on service-aged transformers (more than several years old), which could contain residual gases from previous events.
Some gas generation is expected from normal aging of the transformer insulation. Therefore it is important to differentiate between normal and excessive gassing rates. Normal aging or gas generation varies with transformer design, loading and type of insulating materials. Routinely, general gassing rates for all transformers are used to define abnormal behavior. Specific information for a family of transformers can be used when sufficient dissolved gas-in-oil data are available.
Acetylene is considered to be the most significant gas generated. An enormous amount of energy is required to produce acetylene, which is formed from the breakdown of oil at temperatures in excess of 700°C. Excessively high overheating of the oil will produce the gas in low concentrations; however, higher concentrations are typically symptomatic of sustained arcing, a more serious operational issue that can cause a transformer failure if left unchecked.
DGA is used not only as a diagnostic tool but also to stem apparatus failure. Failure of a large power transformer not only results in the loss of very expensive equipment (costs can exceed one million U.S. dollars) but it can cause significant collateral damage as well. Revenue losses due to customers’ outages may be the least worrisome consequence of a failure.
Replacement of that transformer can take up to a year if the failure is not catastrophic and can result in tremendous revenue losses and fines. If the failure is catastrophic, then additional loses could be realized, such as adjacent transformers, environmental problems from the release of oil, which could be as much as 20,000 gallons, and the resulting fire that must be contained and smothered, all of which are a utility’s worst nightmare.
In order to avoid such a failure, the sample frequency of most large power transformers is between one and three years. However, sampling frequencies will increase as an incipient fault is detected and monitored. Often times sampling frequencies are dictated by insurance requirements, which often stipulate that annual transformer oil analysis must be conducted to ensure continued coverage.
The following examples are extreme but serve to illustrate how the dissolved gas-in-oil test can be used to highlight active problems.
The utility suspected that the magnetic shunt pads attached to the inside of this transformer tank had come loose and were either touching the core (layered steel structure around which the copper winding is installed) or were coming very close. The result was intermittent unintentional core grounding and stray flux causing localized overheating of the oil.
It was suspected that a single line to ground through-fault caused the failure of this transformer. The fault was of enough energy to destroy one of the windings. Acetylene was a predominant gas possibly indicating that arcing may have occurred. The acetyleneto-ethylene ratio indicated the problem could be either high temperature overheating of the oil, arcing or both. The problem was likely in the winding because it involved cellulosic materials, as witnessed by the comparatively high carbon monoxide and carbon dioxide concentrations.
A technician noticed that this transformer was enveloped in a cloud of steam during a rainstorm. This prompted an investigation where it was found that the temperature indicator pegged. It was estimated that the temperature in the unit had been greater than 200°C. It was discovered also that the unit became highly overloaded during some switching functions due to current imbalances in the three phases. The condition probably existed on and off for two years. An internal investigation found the cellulosic paper used as an insulant was brittle and crumbly. Again, the key gas indicators for this condition were carbon monoxide and carbon dioxide.
Oil testing is an important part of a utility’s electric apparatus condition assessment. As these cases illustrate, dissolved gas-in-oil analysis is the most important diagnostic test for detecting a wide range of problems.