Measuring vibrations and the partial discharge of the generator is quite common but gives a rather narrow perspective to the condition of the equipment and does not include the possibility to combine the data with early signals from the environment, i.e. process status, and factors that can have an impact on the vibrations. Today, most of it is guesswork. When we at Fortum eNext first started to develop the service, we aimed at maximising condition monitoring for the whole drive chain in order to gain a larger perspective of the whole process and to be able to form a very precise view of the state of the equipment.
Our remote monitoring service has been developed through a pilot at one of our own combined heat and power (CHP) plants in Suomenoja, Finland, where the last coal-fired unit is scheduled to be decommissioned in 2025. The aim of the ongoing remote monitoring pilot is to guarantee high plant availability for the remaining years and to secure the safe and cost-efficient shutdown of the unit. The decisions on the overhaul and maintenance activities of the unit have been supported by in-depth analysis conducted on the basis of data gathered at the power plant’s turbine island.
Measurement devices installed while the turbine was rotating
The pilot started in spring 2019 with inspections at the site. We were soon able to confirm that it was possible to utilise existing sensors for most of the measurements. For example, the measurement devices for monitoring the partial discharge and the rotor flux of the generator were already in place. Similarly, existing sensors were utilised to measure the shaft vibration.
To start the remote monitoring as soon as possible, a field cabinet was brought next to the turbine and, as a temporary solution, the cabling was routed from the sensors to the cabinet. Finally, access to the process data was created together with viewing access to the operating screens to enable faster remote support if needed. The high-frequency connections had already been tested in Fortum’s lab at the headquarters earlier in the spring.
The tricky part with the pilot project was that the installations were done while the machine was still in operation. However, this is exactly what we wanted to prove: that installing the sensors can be done while the turbine is running; this makes it possible to take the remote monitoring service into use at any time without ramping down production. The implementation of the service is thus very flexible and can be improved gradually. More sensors and measurement devices can be installed at any later stage to complement the service and improve the monitoring level according to exact business needs.
The monitoring level at Suomenoja power plant will eventually be at “Maximum” which is the broadest level of the new service. The maximum package enables comprehensive condition monitoring of both the turbine and generator and supplemented with the relevant process data.
In order to test the scalability of the service, we also installed additional sensors to improve the monitoring level. For example, new sensors for measuring the bearing vibrations and temporary sensors for measuring speed and synchronisation were installed. The permanent measurement devices and sensors for shaft voltage and current monitoring will be installed at Suomenoja during the summer outage, because these are one of the few measurements that cannot be installed to a rotating machine.
Emerging failures spotted before they turn into bigger problems
There are several different solutions available on the market, but, unlike many, remote monitoring by Fortum eNext is based on high-frequency condition monitoring continuously capturing operational data from a local power plant and transferring it to experts for monitoring and analysis without delay. We not only use the automation data available to look back on what has happened, we also bring in our own equipment for measuring the mechanical condition of rotating equipment and the electrical condition of the generator, and we always dig deeper. This enables us to instantly react to abnormal operational situations and anomalies of turbines and generators, spotting emerging failures before they start turning into bigger problems that could lead to an unexpected outage.
All the data is collected to a cloud platform where the analysing tools are installed and to which selected experts have access in order to perform the analysis. The cloud platform also gives us the possibility to share the data with the customer through a dedicated platform. Available data and data access are always carefully defined together with the customer.
Aiming at accurate forecasting with the help of AI
The possibility to develop the product through a pilot project gave us the chance to test if the plans really work in practice. The pilot at the Suomenoja CHP plant further deepened our knowhow and understanding on how to increase plant availability, how to have control over maintenance and investment costs, and how to optimise maintenance and overhaul schedules.
Remote monitoring by Fortum eNext is under continuous development – this is only the beginning. Today, the service enables us to accurately measure the current condition of the equipment. However, the target is to develop a modelling and smart algorithm-based solution that could forecast the future condition of the equipment at any given point in time. This would further enable data-driven plant management and provide a powerful tool for long-term maintenance planning. Needed overhauls could be based on data and not only on the estimated schedule given by the OEM. This, in turn, would enable cost-savings for the operator.