Presented By:
Yuri Monteiro Rossini
Siemens Industry, Inc.
TechCon 2019
Abstract
Power transformers are valuable assets ensuring a continuous supply of power on the electrical network. The cost of a transformer is directly related to its power capacity, which means it makes sense to invest in equipment with proper power ratings, i.e. a transformer that can fulfill the requirements with the minimum cost possible.
The intent is to always keep a transformer well maintained and working under normal circumstances to avoid fast aging and loss of life. However, sometimes it is necessary to overload a transformer during its operation in order to accomplish a temporary requirement, either for a technical or financial reason.
Transformers are sensitive to overloads, as well as operating below nominal power when other issues affecting the transformer can cause overheating conditions. The equilibrium point at which a transformer can be loaded to avoid accelerated aging is the reason this study was developed.
The aim of this presentation is to assess the management of a transformer load and/or overload using an optimized approach based on the current conditions acquired in real-time by an online monitoring system – load, hot spot, and ambient temperature – and the limiting parameters from the equipment design.
Introduction
It is very common among transformer owners to adopt the conservative practice of avoiding overloads due to a lack of knowledge about what can happen with the equipment if the increase of load above nominal ratings is allowed, even for short periods. This situation could be different if the operator had enough information to decide how to load a transformer knowing the effect on paper insulation aging and monitoring its condition during the period of abnormal load. Another common belief is that the accumulated aging should always be kept below 1 p.u. to assure a good operating condition of the transformer. Although it is desirable to have the aging under control and the total loss of life equal to or below the operating hours, it doesn’t guarantee the equipment to be free of failure risks. A more intelligent approach relative to market demand would be to make decisions based on the risks of operating under certain conditions, the real need to operate these assets under specific conditions, and the costs involved with acquisition and maintenance of the assets.
With sufficient data and parameters from transformer manufacturing and testing, an Online Monitoring System can guide the operation of one transformer or even an entire fleet to meet the availability requirement and still have the operational risks under control.
An Innovative Approach to the Load Guide
Since transformers came into use, the management of their load has been a point of interest and study. There are currently several papers and articles from different sources that can be found, and this type of study became part of standards [1] [2] as well. The approach of these sources is static and focused on transformer loss of life. It analyzes the effects and consequences of loading equipment beyond its nameplate ratings, taking into consideration fixed parameters, such as transformer nameplate data and average ambient temperature.
The proposition of this paper is to use all the consolidated studies from the standards and apply them to a dynamic situation, in order to help make real-time decisions and support any unplanned requirement that might force the transformer into an overload condition [3]. The dynamic analysis is able to show some possibilities not predicted with the static approach as the ambient temperature is measured online and can affect the heat exchange between the transformer and the environment. Also, the hot spot temperature curve is continuous and previous conditions are an input to the final temperature and how fast it will be reached.
Transformer Modeling
Transformer online monitoring systems were developed to predict failures and behaviors as there is no accurate way to understand what happens inside a transformer during the equipment operation – unless some critical point is already reached, and the condition presents an outside manifestation. To avoid undesired outages and to reduce the risk of operation, monitoring systems were presented as a tool that can diagnose the transformer based on its current condition [4]. To have this feature implemented it is necessary to develop models that interpret the collected data and create an accurate simulation of the equipment. This kind of modeling must consider details and particularities from manufacturing, testing, and mode of operation from the different types of transformers and reactors, which implies that the most reasonable manner to have this development successful is to base it on specific knowledge and experience.
Among all the models necessary to map the operation of a transformer there is one that serves as a reference for the others and that describes the basic operation of the equipment – the Thermohydraulic Model [5]. It calculates the temperature at the hottest spot of the transformer, which is in the top region of the winding, but it also maps all the heat distribution inside the equipment and the heat exchange due to the insulating fluid circulation. This movement, together with other internal phenomena (such as magnetic flux dissipation and electrical current), helps to explain why the top of the winding is the hottest spot of a transformer, unless there is a failure in another spot of the winding that generates overheating – and that is why it is so important to map the overall behavior of a transformer. In a system with fluid circulation and continuous heat exchange, the thermal modeling is dynamic and the active cooling stage is fundamental to determine how the heat distribution is presented and how the heat exchange with the environment is set.
To reach a more accurate hot spot calculation and, consequently, better Thermohydraulic Modeling, it is crucial to use the bottom oil temperature measurement, i.e. the measurement of oil temperature on or close enough to the cooling system outlet. This is the point where the oil just exchanged heat with the environment and it shows the cooling system efficiency. Also, the bottom oil temperature measurement is used to determine the oil longitudinal gradient, which is the vertical distribution of heat inside the transformer along the oil ducts.
The oil longitudinal gradient (calculated) and the bottom oil temperature (measured) serve as inputs to the Dynamic Overload Guide Model, so the accuracy of the measurements and calculations are of extreme importance to the system. Other measurements used to calculate the maximum load availability are the ambient temperature, transformer load, the top oil temperature, and the cooling status. All values must be online and the cooling stage status must be captured from the cooling system itself, instead of the cooling system controller, to avoid any miscalculation due to failures on system activation. The ambient temperature is the environment temperature at which the cooling system heat exchange process takes place, so for a water-cooled transformer (for example ODWF) the ambient temperature is the water measurement, which must be taken on an inlet spot before the exchange region.
Applying Online Monitoring Systems to Load Management
The question then to transformer owners is how to use online data from monitoring systems to improve the load management of the assets. The Dynamic Overload Guide Model uses the hot spot temperature calculation and provides the emergency loading value that the transformer can operate at with the maximum cooling capability and with no cooling capability, in case there is any problem with the cooling system. It allows the user to plan overloads even in critical conditions, where the cooling system is not available. The model also shows that the load limit is not always the nominal rating. As the calculation is dynamic, the model indicates in real-time the load limit that is dependent on ambient conditions and the current load of the transformer. If the system indicates a load limit and the user must exceed this limit for any reason, this will directly influence future performance.
The model always keeps the limit hot spot temperature as the guide parameter to provide the load limit, because this will keep the aging factor to a maximum of 1 p.u. However, if any other input indicates a lower hot spot temperature limit, this is taken into consideration to update the Dynamic Overload Guide Model outputs. Examples are water bubble formation temperature that can be affected by sudden change in ambient temperature, moisture level in the oil and insulation paper, and many other factors. If the moisture level is available and the system indicates a critical temperature for bubble formation below the hot spot limit that can be reached by transformer overload, the system will perform a cross-check and adjust the available load limit to a level where the hot spot temperature doesn’t reach risk levels due to a different factor, in this example the bubble formation. As the system always indicates the predicted hot spot temperature with the use of the indicated load, it is clear that the limiting temperature in some circumstances is due to some input other than equipment aging. In the examples, there are no external limiting factors since the temperature shown is the maximum hot spot rise temperature or the closest it can be, as a discrete increase of the load can lead to exceeding the hot spot temperature limit.
Using all the features available with this tool, the user can respond quickly to load changes. These changes in load requirements can occur due to failure of any parallel equipment, grid flow transactions, and residential or industrial demand. In any of the cases, the decision must be made quickly and the tool presented provides powerful real-time information for doing that. From a different perspective, the monitoring system also serves to support maintenance or even replacement planning, since a full analysis of the historic load profile can be compared with the condition of the transformers during those critical periods. Furthermore, with the use of several types of sensors, the monitoring system can also provide a complete scenario of the transformer condition and trends, which can be used as another input for maintenance planning.
Conclusions
The content presented in this paper shows how to use transformer online data to evaluate the available options for load management. It goes beyond the standard definitions and applies all knowledge on the topic to a dynamic approach, where decisions are based on the aging evaluation in real-time from the online acquisition and interpretation of data, as well as the current load demand from the power grid. By providing the availability and also a condition evaluation of the transformer, the system allows decisions to be based on true data and avoid the conservative scenario due to lack of knowledge of the equipment status.
An interesting approach that can be developed by the user, or even used by consultants, is the comparison evaluation between the utilization level of the transformer and the remuneration received for its overload. This study is most applicable to critical situations where the load increases suddenly and a quick decision must be made. With enough inputs, it is possible to decide if even a short-term accelerated aging event is worth the financial benefit from allowing the overload.
References
- IEEE C 57.91 – “Guide for Loading Mineral-Oil-Immersed Transformers and Step Voltage Regulators”, 7 March 2012.
- “Advances in DGA Interpretation”. Cigre, 2019. TB77IEC 60076-7 – “Power Transformers – Part 7: Loading Guide for Mineral Oil-Immersed Power Transformers”, Edition 2.0, January 2018.1.
- “Permissible Loading of Oil-Immersed Transformers and Regulators”, Facilities Instructions, Standards and Techniques, Volume 1-5, April 1991.
- “Power Transformer Condition Monitoring and Diagnosis – IEEE”, The Institution of Engineering and Technology, 2018.
- “Transformer Monitoring and Diagnostic System (TDCM) – Thermal Model Description”, Siemens AG, 2014.
