The reliability route to less downtime
03 December 2020
With lockdown and social distancing measures still impacting industrial sites and the need to be competitive in an increasingly global market becoming ever more vital, one thing that businesses cannot afford to have is unnecessary downtime. The recent IP&E Live conference on Industrial Maintenance, sponsored by Omicron and Megger, took an in-depth look at reliability best practices
The first presenter on the day was Floiran Fink from Austria-based Omicron, who spoke about critical asset monitoring in industrial power grids.
“I think it’s more or less clear, when you have a power outage, you will probably have production interruption and this will lead to money loss. So it’s always the goal for the person responsible for the power grid to avoid such power outages, said Fink
Fink explained a typical industrial power grid, asserting the importance of identifying the critical components that can lead to an outage, so it can be avoided. In his experience, talking to industrial customers, he has found that many intuitively identify potentially problematical assets or situations and it is good to verify this via risk assessment.
In one example an outage occurred due to protection miscoordination. Expanding continuously, a company in the paper industry, expanded its grid to match accordingly on a project by project basis and new machines meant expansion of the power grid. Power grid components such as switchgears, cables and transformers to expand the power grid were installed by project companies.
Protection coordination was carried out only for assets related to a particular project - and different projects were carried out by different companies. Consequently, a short circuit at a transformer led to complete standstill of the whole production. As a result, it was realised that the overall protection coordination and control should also be viewed as ‘critical components’.
In another example, a company experienced a ground fault on a medium voltage cable, which melted the insulation. High currents were detected on the copper shields, caused by induced currents from parallel cables. A solution came in the form of a new measurement to determine the potential shield current, before commissioning, to enable countermeasures such as cross bonding to be implemented and ensure the quality of the power grid.
A further case quoted by Fink involved a new power plant at an industrial site which incorporated chlorine electrolysis being fed through 2km of medium volition cable. In this case the generator, transformer, cable and chlorine electrolysis created an oscillating circuit. Resulting harmonics and overvoltages created partial discharges, leading to damage to cable dealings, cables and transformers. To resolve the problem, several interruptions to production were required so that repairs could take place.
So the issue does not occur again, harmonics and partial discharge measurements have been taken to identify damaged components, and a harmonic filter was installed in close proximity to the chlorine electrolysis.
Analysing the grid
When it comes to analysing a power grid, there are two different types of assets. With the first type of asset, failure will cause an outage of production to stop; failure of the second type will not result in an outage of production - the goal, said Fink, is to reduce the number of assets that will lead directly to a production outage in the event of failure. However, from a financial point of view, often this isn’t possible. Plus, on occasion, it is difficult to identify whether an asset will fall into this category until it actually fails - this can be negated by carrying a risk assessment to determine which category an asset falls into and evaluating the possibility of failure.
For assets that cause production outages on failure, it is important to implement an effective testing strategy, for example, off-line testing and maintenance, or online monitoring. It is also good to look at potential scenarios and the knock-on that they could have on the power grid if a component fails - something that Finks says is often not particularly obvious. In order to achieve this, it is beneficial to have good relationships with consultants, service providers and project companies.
For the end-users, gaining an overview of the assets and their testing status, in general, is advantageous, particularly if one of the partners changes, and also in view of any legal requirements.
Of course, new technologies bring advantages, but they also create new or different tasks, for example with digitalisation much more data is available for testing and measurement, which can lead to different servicing requirements. In such situations, Fink says, it’s not simply a case of buying a new test solution, you need to plan in advance and involve service engineers, consultant and service providers in any change of process.
Critical motor testing
The second speaker of the day was Mike Herring, regional sales manager at Megger, who covered electrical testing of critical motors. He opened with the premise that: “While vibration analysis and other technologies are commonly used as part of a predictive maintenance programme, the motor itself and its winding from an electrical point of view are often given little or no consideration.”
Electric motors are found across all industries and are often critical to plant operation. Should a critical motor fail, it brings about unplanned downtime, resulting in lost production and therefore revenue. There is also the cost of motor repair or replacement to be considered, potential loss of reputation, and in some cases involving safety or environmental issues there can be fines and litigation.
Herring referenced a GE study showing that for motors up to 4kV, the most common cause of failure is the bearing; vibration analysis is widely used to mitigate this. However, the next most common cause of failure is stator related; to capture these developing faults different technologies are better suited. For motors rated above 4kV, 66% of failures are stator related, so there is a need for comprehensive electrical testing to detect various types of electrical failures that can occur. The electrical tests complement the mechanical tests to give better coverage of all the failure modes possible.
Dynamic testing is used to analyse the running motor by measuring the voltage and currents of the three phases - this is not just a motor test, it monitors the whole motor stream system: the power supply feeding the motor, the motor itself, and the driven load.
Applicable to both low- and high-voltage motors and generators, static testing is a multi-test approach to detect motor circuit problems and insulation degradation. It can be used for fault finding, predictive maintenance and quality control. Results should be trended when used as part of a predictive maintenance programme.
Using safe working practices, static testing typically occurs in the motor control cabinet and includes the power cable and the motor. If a fault is detected, the motor then has to be tested without the cable in order to detect the fault location.
One of the main aims of static testing is to establish the integrity of the insulation. The most commonly used tests are the Megohm/IR test and phase to phase winding resistance measurement; for high voltage motors a PI test is often added. While in many cases these are the only tests applied, they are not sufficient to detect all types of developing faults, said Herring. These tests are typically performed at or below the motor’s rated voltage, which means they are incapable of detecting a weakening of insulation that may be developing within the motor.
Static motor testing is performed using a multi-function tester, which performs a series of low voltage tests and then gradually proceeds to high voltage tests. LV tests measure specific electrical parameters at or below main plate voltage. The aim is to detect a change in the electrical circuit properties, such as resistive unbalance. High voltage tests use voltage level similar to those the motor encounters in normal daily operation. This helps to determine the integrity and likely reliability of the insulation system.
Test Voltages are covered by various standard bodies such as IEC, IEEE, NEMA and ESA.
Herring went on to detail high voltage testing of a 400V motor. The insulation’s dielectric strength when new is around 6kV AC, although the value does vary depending on the specification of the wire. The paper slot liner - the barrier placed in the slots that separates the winding from the frame - is typically good for around 20kV. If a 2kV test is performed, this is way below the expected dielectric strength of the insulation system, so if a fault is detected it indicates that the insulation has already degraded. It could also mean that the motor is contaminated, but a fault will not be caused by carrying out a high voltage test.
Testing above the line voltage is carried out because motors are regularly subject to voltage spikes from contractors during start-up and shut down, and also from inverter drives. These switching surges can be as high as 2kV on a 400V motor, so to understand if a critical will continue to be reliable it is necessary to test at similar voltage levels to the surges. This is done via a low energy test in a controlled manner, which will give an indication of any weakening in the insulation.
“All tests have some limitations and it’s important to understand what those limitations are, and where possible to use additional tests to give additional information with which to make decisions,” said Herring.
Surge testing is widely used by motor manufactures and repairers to detect copper to copper winding issues including weak insulation. Paschen's Law states that in order for voltage to jump across two conductors, a minimum potential difference of 325V is required. When a motor starts it is often subject to fast switching surges, which creates a potential difference between the turns of the winding. If the dielectric strength of the insulation has reduced down to the level of those switching surges arcing can occur, accelerating insulation damage and causing premature motor failure.
The surge test injects fast rise time pulses into the winding and creates the required run to turn potential difference, checking for any weakness in the insulation. This is done with very low energy, for example a 4kV tester has pulse energy of less than one joule.
Comparing the surge test to the megohm test, with the megohm test a dc voltage is applied to all the turns in the winding. This means the potential difference between turns is zero, and according to Parschen’s Law, a turn to turn weakness will be undetected within one phase. It can perhaps be used phase to phase, but only if the neutral points can be opened, and a high enough voltage is used, said Herring.
While static testing looks the the motor circuit and insulation condition, dynamic testing aids understanding of the stresses places on the motor during normal operation, providing vital information on the entire motor driven system - ie. the power supply feeding the motor, the motor itself and the driven load. It also gives an indication if the power supply and operation of the motor is likely to effect the insulation integrity due to heat caused by power issues or load level.
Dynamic motor analysis is primarily used to monitor motors on site, operating in the normal working environment. The voltage and current is measured from all three pahses and is used to caluuclute a host of important parameters, effetely using the motor as a sensor.
Motor current signature analysis (MSCA) is similar, but by adding volt measurement it is also possible to asset the power quality, which is vital becasue power quality issues will result in increased winding temperature and heat, said Herring, is the main enemy of insulation. By looking at the current and torque spectra, it is possible to understand the condition of the driven load.
Dynamic motor testing is carried out from the control cabinet, so there is no need to see the motor, making it particularly useful for applications such as submersible pumps and motors in restricted areas, for example radiation zones), where vibration sensors are often not fitted and there is no other way to test the motors.
In order to carry out dynamic testing it is necessary to access the voltage and current signals. Maximum voltage measurement is 100 volts, so for high voltage motors measurement has to be made through the voltage and current transformers within the controls panel. This type of connection requires the panel to be open, so the motor needs to be stopped and isolated. If it is not possible to stop the motor, an EP1000 unit can be permanently mounted to measure voltage and current signals - a tester can then be plugged in to collect eh measurement data without having to open the panel door and exposing the engineer to high voltages.
The online seminar also included presentations by representatives from the United Kingdom Lubricants Association (UKLA) and the HSE, the Manufacturing Technologies Association (MTA), the Lifting Equipment Engineers Association (LEEA), and British Compressed Air Society (BCAS), which will be covered in later issues of the magazine.
For more details on the topics covered here, you can view IP&E on-demand by registering at https://bit.ly/36BioZJ