- I am Mark Frogley. I represent Acoem. We're a vendor of condition monitoring tools and technologies, as well as reliability tools and provide training. And we're gonna talk today about wireless wired and portable strategies that you may want to employ as part of your reliability programs in your plants.
So the first thing to really accept is the law of machines. Every machine that gets installed will wear out or be forced into failure. The trick is that we need to detect that degradation in machine performance at the right time so that we can make a plan for repair, an intervention plan, or a mitigation plan to effectively make that machine run longer. All of the tools that are now available to us as part of the wireless wired strategies of vibration monitoring will help us do that.
What we're trying to achieve is to contribute to organizational knowledge. So we all talked about the Internet of Things, AI, machine learning. We want to utilize a wide sensor array and different methods of gathering machine information so that we can harvest that data and turn it into organizational learning as we're on the journey of proactively identifying defects, degradation, identifying the failure modes that are common to our organizations, being able to address those through training, tools, practices to make them more effective. At the end of the day, we want to make sure that our cost and effort is efficiently employed so that we can get the maximum effectiveness out of them. So too much data is a bad thing if you're not using that data.
Data rich, knowledge poor. Choose your monitoring strategy. And I think one of the most important things is to understand the criticality of the machines, understanding the components that are in the machines or assets and understanding how you want to work to get meaningful data. Now, all of the technologies are very effective and deployed correctly, you will get the benefit from them. We talk about machine criticality. This is an example of an OEE based criticality matrix for manufacturing.
And you can see six parameters that they're quantifying at the level of high, moderate, and low. And you want to be able to quantify in each of these boxes under each column heading and share that with your stakeholders so that you can develop consensus. But they have to be measurable, not just random. I think, or my opinion, you need to construct these criticality matrix against the assets or machines that are in your facility and the business plan for production. 'Cause after all, at the end of the day, we all get paid from product put out the door and sold.
The next step. Once we've effectively defined what are our important pieces of machinery, is to define their failure modes and what's the most effective failure detection technology. Different machines are going to fail at different rates depending on what they're doing and what their design is. It could be hours, it could be days, it could be weeks, it could be months. Typically, you're gonna want to sample the machines at four to six times the rate of failure. Today we're not going to talk about protection systems, which will typically need to be operated or triggered within a revolution of a failure to prevent catastrophic damage to equipment and personnel and environment.
What we are gonna talk about is how we employ vibration. So you can see from the chart that we've looked at this baghouse fan and it's got different failure modes. It wears out, it can have a performance failure, and of course, we've got a quality box there that we always need to pay attention to that.
The technology that we're gonna be talking about today is vibration, and we're gonna talk about the different sensors and the different capabilities and requirements for those. What we'll be speaking about will be MEMS sensors, the micro electrical mechanical system sensors that are now widely employed in industry, but a fairly recent trend for the reliability professional. We're gonna talk about wireless piezoelectric sensors, which are also available, wired piezoelectric sensors for surveillance systems like an MVX, route-based systems that can use wireless as well as wired transducers or remote transducers switch boxes.
Traditionally we've used accelerometers in our industry for measuring vibration, and these are reliable and effective, they come at different cost levels. Depending on the quality of the transducer and what your requirements are, they can be used with portable instruments or online surveillance instruments or through some wireless systems. The MEMS sensor, as I said, was a relatively new sensor in sensor technology into the reliability realm.
They're becoming increasingly deployed in many plants. They are reliable, they are effective, but they've got a different construction than our typical piezoelectric sensor. And we're gonna talk about that in just a moment. So on the left, you'll see a piezoelectric sensor or an IEPE/ICP accelerometer.
And on the right you'll see a typical construction of a MEMS sensor. And you can see that the IEPE accelerometer uses a piezoelectric crystal, which when depressed, will create a small charge that's measurable as a vibration, that changing charge is being measured by vibration. They're typically coupled with some sort of electrical transmitter to be able to get the signal improved and back to the data collection device. MEMS sensors are using a different technology in terms of capacitive discharge and measuring that change in the capacitive distance between the mechanical pieces.
What's important to recognize when you're making your selection is that both work well, but they have different capabilities. And what we're seeing here is with just some typical capabilities. One of the first things that you're gonna note is that the frequency range of a piezoelectric transducer is wider than the piezoelectric range for a MEMS sensor.
What this means is that you'll be able to detect fault earlier with a piezoelectric transducer than you can typically do so with a MEMS sensor. The other point to look at is the temperature range. Temperature is gonna play an important factor in where you can and can't put our sensors. Dynamic range is another important factor, and that's the difference between the smallest signal it can measure well at the same time, the largest signal. So again, another important parameter. A lot of discussion has also been about measurement life.
Of course, wireless transducers, MEMS sensors require batteries. Those batteries will degrade and they'll degrade at different rates and need to be changed depending on how frequently you're measuring vibration and transmitting the information over radio frequency or wirelessly. Lines of resolution, the last piece is also important, and that's important to be able to differentiate the more complex vibration signatures that may come from complex machinery like gear boxes, machine tool spindles, that kind of thing. So you need to be aware of what you're actually trying to detect and when you're trying to detect that, make a plan, think it through, and be mindful. Collection strategies and techniques. Wireless sensors utilize radio frequency, which means that they're going to be based on line of sight.
Now there'll be opportunities with many systems where you can put repeaters in, which can take that signal and then transmit it further, matrix systems or mesh systems. So there's lots of different things that you've gotta consider. But if the end of the day, it is radio frequency, it is line of sight and you can get technology to help you do that.
Now when we're talking about wireless, it's typically cheaper on a per sensor basis, but it will require maintenance because batteries need to be changed periodically. Wireless systems that utilize piezoelectric are often more expensive for a fixed system as opposed to a portable system, which is wired but more affordable, if you will, because you can get more coverage from them. But again, you have to have access to the machines. So when you're deploying a wireless or a wired system, you're gonna be looking at machines that require monitoring that may not be accessible and are in the right environment for the transducer that you're selecting and the technology you're selecting and may perhaps require a more frequent sampling or vibration collection. Today we're lucky, we've advanced to be able to include all the sensor technologies into one platform.
So again, coming back down to being able to put people, process and technology together to develop a mindful strategy you can synergize, the effort of the people, the process and the technology to create organizational knowledge around the machine health and strategies to improve the installation and life of that equipment. What we're trying to do is not just collect data, we need to plan when we see degradation so that we don't interrupt production. But along with that journey, we also want to understand the fault modes and how machines are failing so that we can change our maintenance and operation practices to make those machines run with less effort longer. So it's all about efficiency.
We talk about the PF curve, so the prevention failure curve, and there are four aspects to this curve. We're gonna talk about this a little bit. So design, of course, from a monitoring standpoint, the design and engineering procurement is really about selecting the proper machine for the proper job. Did we get a pump that is gonna operate on its curve properly? So that's something that comes upstream. And typically the reliability professionals will not get involved in that until it's handed over for installation. At installation, the reliability people want to make sure that it's installed and commissioned in a quality way.
That means we've got our alignment dialed in, we've got our balance, we've made sure that the process parameters and what it was designed to do are within spec and that we're not seeing any deviation. Once installation is complete, we're gonna put it in to commission it and then into operation. At the point that it goes into operation, we're gonna start to monitor it and we're gonna want to know when degradation occurs, ideally as early as possible, but that's not always something that we can do. We're gonna start to the left when ultrasonics are available to us and move down that chart. And then finally, once we get down into functional failure, which is basically when the machine fails or craters and stops, we're too late. So again, when we talk about the different technologies and wired wireless and portable systems, we need to pay attention to the PF curve and where we want to detect degradation and how long we have until we need to plan for an action.
So a high frequency sensor, we want to predict and we want to be proactive. So often with a vibration sensor, a piezoelectric sensor, we can detect lubrication problems, apply lubrication to the machine before damage occurs, and effectively make that machine run longer. In the latter stages of degradation of a bearing down around the medium where we see the MEMS sensors could apply, we could be, you know, short months or weeks before failure. In that case, we may be too far along in the degradation to prolong the life of it, although we can to some extent, but we can certainly plan for replacement and repair at that point.
Once we get down into the low frequency sensors, we're basically what we call in the industry, an earthquake switch, which will shut the machine down or throw up an alarm that, okay, now you've absolutely gotta do something. Shut down, fix, move over to your spare. Hopefully you were monitoring your spare and carry on. So, we at Acoem, we run a remote diagnostic center and we took a sample out of our database of 25,000 analysis that we've done and random sample. And we've seen the big four faults and we see a lot of cases of lubrication and we're identifying this in the early stages, and our customers are able to proactively go in and lubricate those bearings or correct their lubrication practices. We're also seeing a lot of misalignment that goes in from improper commissioning often.
Sometimes base defects creep in, and we're gonna start to see misalignment. Bearing defects, as bearings wear, they will produce faults and degradation that we can detect from the early stages to the latter stages as we discussed on the PF curve. And then we're gonna see a lot of balance and resonance issues. So a lot of that is also taken care of proactively in the commissioning phase.
Your programs are going to see the same type of fault distribution. What you do about them and how you address them through planning and through proactive maintenance practices are gonna be how you gain success for your organization. Plant process. So this is a fairly simple diagram just to depict an idea or a concept of where we would deploy a MEMS sensor, a wired piezoelectric surveillance system, a wired or wireless piezoelectric system, or a portable system. So the first, we've got our exhaust fan up in the top right corner, and in this case we're gonna put a MEMS sensor on it. We know that that machine's gonna wear out at some point.
It's inaccessible, it's up on a landing, and we want to get data fairly frequently. When we are notified in the latter stages of degradation that something's going on, we may choose to increase the sampling rate so that we can monitor the rate of progression of the failure, and in that way, maybe nurse it into a planned outage or get maintenance done at the right time. So again, this fan is up on a platform, it's not accessible, it is critical to the facility, but the failure mode is such that we can detect degradation, monitor that degradation, and then plan for maintenance.
The next system we're gonna deploy is a portable system. So we've got our little circulating pumps down on the floor here, very accessible. We all know that pumps, when they're installed correctly and operated correctly, run with a high degree of reliability. So we can sample on a monthly basis, we'll walk around and check those. One of the advantage of the portable collection is that when we're there, we can also note the suction and discharge pressures to make sure that the pump is actually operating on curve. Although we know that the vibration technology is gonna tell us if they're cavitating the pump, it's always nice to have eyes and ears around these rotating pieces of equipment to help the operators run them or optimally run them.
On our machine in the center, which is critical, a high value machine, big replacement value, not that accessible, we're going to use a surveillance system with wired transducers. And the reason for that is we want to catch that early stage defect so that we can take preventative measures or proactive measures to extend the life of the machine. In this case, the parts for that machine have a long lead time. So the earlier we know we're gonna get some kind of degradation, the sooner we can plan for replacement and order the correct parts for it. And finally, we've got our pumps, our well pumps out here in the middle of nowhere, and we all have those. We don't like walking out there in cold weather or any weather, it's a long way from the plant, but they are critical to production.
And again, we're gonna need to use a wireless system because we don't have the ability to run a cable. And so we're gonna put a piezoelectric wireless system on these and we're gonna be able to detect early fault degradation in these machines. And when we detect something, we'll be able to sample more frequently and as necessary, we'll be able to order parts in enough lead time to make it for the outage.
So criticality is another way of looking at this. So we're talking in in sort of more basic terms that the criticality of the machine may be an easy way of defining your technology for data collection. So we see on a monthly basis, traditionally we use portable, you see on the left on demand, we use devices to ensure quality installation at commissioning, maybe some monitoring and spot checking, but not typically for continuous or route based or program management. Daily or weekly, we're gonna get into MEMS sensors and wireless piezoelectric systems. And then semicontinuous or more detailed monitoring would be surveillance systems on the right side.
And then as I discussed earlier, turbo machinery will typically have some kind of protection device on them, which will shut down the machine if a fault occurs, typically within less than seconds within shaft rotations, but often we can collect condition monitoring information from those systems by collecting it either with a portable device or with a surveillance system off the back of the rack. So just to give a flavor for the size of some of these sensors, you can see that on the right side of our pump set, we've got some wireless transducers using piezoelectric sensors. They will connect to a gateway, and this is a fairly standard mechanism for using wireless piezoelectric systems, get to the gateway, then the gateway retransmit either directly over the land or again, repeats onto a wireless system so that you can pick it up on the cloud directly. On the top we've got a MEMS sensor and you can see the size profile of our MEMS sensor on the vertical position of the drive end of that motor. A little bit smaller, a little bit shorter, but still there's some space there. And then on the left, as we move down through the drive train to the drive end bearing of the pump, we see a wireless transducer.
This is a three direction transducer that communicates wirelessly with a portable data collector. And then on the bottom we see a couple of accelerometers, which are traditional and they're in the horizontal direction, and in this case, depicted as being connected to a surveillance system like the MVX. Now, these transducers can also be utilized with the portable systems that we all have to do collection on a monthly or weekly basis. What's important to look at is the size profile, because that's another factor that is going to influence where you put your transducers or where you can put your transducers. We all know that you want to get as close as possible to the bearing, and you want to have a clean path for the vibration to transmit from the bearing to the sensor that's gonna give you the best vibration that you can. You can place it in other orientations and in other parts of the machine, but you will degrade the vibration and signal.
And finally, as we talked about Industry 4.0, machine learning, the ecosystem of all of this data, bringing it together so that you can be effective with improving the reliability of your machines from commissioning, improving your productivity output by being able to plan maintenance to occur when production is available for it, and to go back to organizational learning so that you can measure and quantify what has been happening in the failure modes of your machinery so that you can adopt new maintenance practices either through training tools or new strategies to condition monitoring. We are fortunate now that it's very easy for our industry to effectively expand our programs and measure the bottom line returns.
So in summary, what we're trying to do, achieve organizational knowledge, we want to plan proactively so that we can not influence production. We want to optimize our machine operation so we extend the life of it, and we want to do that at the best cost, the most effective way we possibly can. Trying to get all of these technologies to work together in a hybrid system, in an IoT environment, in a cloud-based environment.
2024-02-09