Reliability and good component life is no accident, explains Angus Lay.
One question I often hear, is “why did Engine A achieve its target life while Engine B only achieved 85% of its target life, and was removed due to excessive blowby and poor condition?” Both engines were fitted to the same make and model of machine operating at the same site, both had the same operating conditions and maintenance strategies, yet Engine B fell well short of its target life. Review of the condition-monitoring data showed oil samples were normal until the engine developed signs of blow-by and the downloads from the machine showed no OEM alarms that would have contributed to the early deterioration. Sounds familiar, doesn’t it? There had to be some sort of underlying condition that reduced the component life but what was it?
A common misconception is that because there are no OEM alarms, the machine is running normally and in a healthy way. Let’s unpack the underlying causes for reduced component life. All components, whether it be an engine, transmission or hydraulic system, have a normal operating range for the different parameters, from temperature, boost pressure, oil pressure, etc. When the component is operated within these normal operating ranges, the stress on the component is at its lowest. However, when operated outside these normal operating ranges, whether it be too hot or too cold, too low or too high a pressure, the stresses on the component increase. In the case of increased operating temps, it has a number of detrimental side effects, first as the temperature of the engine increases the oil viscosity reduces. Second, we have thermal expansion as the temperature increases, the internal components expand, affecting the internal clearance. Combined, these two conditions result in increased wear rates, and while it may only be a slight increase, if left unchecked over an extended period of time, these increased wear rates rob the component of potential life.
Now back to the OEM alarms: to put it simply, the purpose of these OEM alarms is not to ensure the machine operates the normal low stress range; it is to prevent immediate severe damage and/or catastrophic failure from occurring. Quite often, the gap between the OEM alarm and the machines normal operating range can be quite large. For example, the OEM alarm triggers at 105°C, yet the optimal normal operating range for the engine in question is between 82-87°C. From the engines upper normal range of 87°C to the OEM alarm trigger, is a range of 18°C.
If the cooling system is partially compromised due to the radiator plugging with dirt or mud, or one or more of the thermostats is malfunctioning and no longer opening fully, this can result in the engine operating well above its normal operating range but not trigger any overheating alarms, having not exceeded the 105°C threshold. These conditions can develop over months sometimes years, without any indication of a developing issue until the 105°C threshold is exceeded. The variance of the engine temperature from idling to high rpm high load can also be considerable and can exceed 15°C during a cycle, this constant expanding and contracting, has a follow-on effect on exhaust manifolds, turbos and intercoolers and can result in cracked intercooler cores, exhaust leaks and reduced turbo life.
Real-world data
In Chart 1 the data displayed shows 60 days engine coolant temperature data during normal operation of the machine, data has been filtered to remove periods when the machine is warming back up after a shut down.
Machine 1 shows a minimum coolant temperature of 80°C and a maximum coolant temp under load of 91°C, with the average operating temp being 85°C.
Machine 2 shows a similar minimum coolant temp of 81°C however, the maximum coolant temperature recorded during the same period of time was 104°C, with the average coolant temperature being recorded at 91°C.
Chart 2 is the same engine coolant data as chart 1 however, chart 2 shows the coolant temperature delta between the minimum and maximum operating temperatures. Machine 1 shows a maximum engine coolant temperature delta of 11°C for the 60-day period. Machine 2 Shows a maximum engine coolant temperature delta of 22°C for the same 60-day period.
So how do you detect these issues that are causing cumulative damage and robbing you of component life but are going undetected? First, you need a device such as Monico’s mCore SDR (i.e., an industrial data capture device) that has the capability of reading from more than one data source and potentially multiple different protocols simultaneously. For example, on Caterpillar machines the data required may be distributed across both the Cat Data Link (CDL) and j1939 networks however, on a Komatsu electric drive haul truck depending on the model would require up to four separate connections including CAN, RS282 and RS485 to access all the relevant data from the machine. You then need a device that also has multiple options to offload the data, either via the site radio/mesh networks, local cellular network provider or satellite if required. Just be careful though, as there are some hardware providers that believe you should pay to access the data from your machine, looking to lock you into costly monthly subscription for you to be able access your own data! This device will also need to have the ability to store and cache data when out of range of the network, then pick-up again and transmit this cached data along with the new, when the machine is back in range of the network. You want this device to also be OEM-agnostic, so you don’t end up with multiple different devices and service providers across your fleet just to get the information from all of your equipment. Most mine operations tend to have a variety of different make/models of machines, so this flexibility is critical.
Once you have access to your data and its being stored, you then need a platform that can display and analyse your data in real-time, to alert you to when there are developing issues on your equipment, such as Monico’s mGuard MachineWatch. This platform is built on the Aveva Pi system but also comes preloaded with all the analytics to identify these issues at the early onset. The power and flexibility of a device such as mCoreSDR streaming real-time machine telemetry data to a platform for storing, analysing, visualising and alerting, enables the identification of these outliers, often days or months before the condition would typically result in a breakdown event, and typically well before any long-term permanent damage can be sustained. This early detection also offers time to verify the condition is true and valid, diagnose the root cause correctly, order the required parts and schedule the repairs to be carried out in a planned state during a normal PM service, resulting in little or no additional downtime and ensuring the component life is maximised to its full potential. As you might deduce, reliability and good component life is no accident, and with readily available technology and analytics you can increase the life of your machines.
Angus Lay is with Monico Monitoring.
