BY Atonix Digital Team
This conversation was captured in the webinar: EthosEnergy Uses SaaS Analytics to Better Understand Asset Performance, Prevent Costly Failures and Increase ROI. It was edited for clarity and conciseness. This conversation broadly reflects the most commonly asked questions from Plant Managers and Engineers considering an Asset Performance Management (APM) solution.
EthosEnergy is an independent provider of services and solutions to power, oil & gas, and industrial companies. EthosEnergy relies on two SaaS products from Atonix Digital - Performance Analysis and Monitoring & Diagnostics - to improve understanding of asset performance and prevent costly failures. Built on the secure ASSET360® cloud platform, these SaaS products evolved as internal resources for Black & Veatch, and are infused with over 30 years of power plant operations experience.
Q. Why did Ethos move to ASSET360?
That’s a great question. There are three main reasons.
1. First and foremost, the series of platforms we were using were both too slow and not feature-rich enough.
2. The industry is shifting toward advanced pattern recognition and machine learning, and it was time to get on board with this cutting-edge technology.
3. We were using a more diverse series of software to do our monitoring and diagnostics before, and the management, support and amount of resources we were pouring into supporting those systems was ever-growing. We needed to transition to a system with one point of contact with APM and cloud deployment. AtonixDigital fit all those bills while providing future development for us.
Q. What is the range of reduced operations & maintenance costs?
It comes into two main silos.
1. Direct reduction of maintenance costs. With ASSET360, you change your approach from a direct forced outage to a planned outage That’s an immediate cost saving, and it comes from the early warning.
2. The more critical and larger cost savings come from reducing lost costs. When you move from a forced outage to a planned operation, you never have to go offline. To quantify those savings is harder but undoubtedly larger. We do that by establishing calculation methods within the IssuesManagement module of ASSET360. We’re able to calculate expected loss generation, downtime, foot fees and fines, and the hard costs of repairing a pump or replacing the filter— material costs.
Q. Could you address typical challenges you face when implementing ASSET360?
Historically, there has been a lag period between the initial cost of investment to the realization of the gains you receive. So, from the onset of implementing ASSET360, we captured low-hanging fruit.
We investigate how equipment is performing and might perform in the future. Part of the rollout is determining new math calculations with some machine learning in the background and leveraging existing data. We filter through the data to see what’s critical and required for operations. In a sense, that’s how we perform the rollout.
Q. How long does a typical rollout take?
The typical rollout to “go” is about 6-8 weeks. It can be faster if it’s a smaller system. But even large-scale deployments at a large facility can still be done in that 1-2-month time frame. It’s a streamlined process.
Q. What is the scope of equipment that can be monitored?
That’s a great question. If there’s an analog signal in your plant, we want to monitor it in ASSET360.
We monitor full combine cycle power plants—every piece of major equipment and some smaller pieces of equipment as well. If the data is available in your historian, we monitor it.
Q. The most challenging part is collecting data and formatting it in a usable way. Is that the major part of what you do during the rollout?
Absolutely. The value we’re getting out of ASSET360 and any data analytics platform is only as good as the data coming in. We do a data audit to make sure everything is coming into the historian at an effective sample rate. A lot of times the problem isn’t with the instrumentation, but with its integration into the historian. Due to storage restraints and other plant historian systems, we like to provide suggestions and optimization there as well.
Q. Do you recommend completely replacing time-based maintenance with predictive maintenance?
In general, no, mostly because the size and complexity of assets that we’re working with is quite large.
If I was confident that every discernible aspect of equipment was properly instrumented and modeled, my answer might be yes. However, our main focus is on enhancing the ability to prevent issues before they happen. We want to do more maintenance so that fewer big problems occur.
Q. How is ASSET360 updated after an outage, upgrade or other equipment enhancement?
From a user perspective, it is extremely intuitive to update a model within ASSET360. We can segment out when the historian isn’t working or when extraneous data occurs regarding a particular instrument. Then we can remove it while maintaining visibility.
We’re also able to rebuild the model following a maintenance action or an outage as soon as the data is available. As long as there’s runtime on the units, the models can be rebuilt within seconds to minutes, as opposed to waiting months.
Q: How does ASSET360 adapt to a facility that operates at various turndown levels?
This is one of the major benefits of moving toward advanced pattern recognition and machine learning models, as opposed to what I like to call “old, stationary or dumb” models. Smart models are able to adapt to the two major things that will change the signature of a signal: 1) Environmental conditions—seasonal changes, weather changes, as well as fuel changes; and 2) The operational profile of the site.
ASSET360 captures the whole gamut of operations. It can capture what it believes the model should be and give us an output of where things deviate or become abnormal. That is something that was just not possible with stationary alarm limits or dumb models of the past.
Q. Does the software reduce alarms by filtering them by severity?
ASSET360 filters alerts or alarms by priority based on user-specified parameters.
These do not affect the alarm levels in the site control system. There’s an added benefit that if we can catch something early, alarms never show up in the system. So you’re alleviating alarms that way, but it’s not changing the site’s control system alarms.
Q. Does Atonix Digital have experience with the process industry?
While our products have been widely deployed in the generation industry, we’re rapidly deploying across various industrial processes. We’ve implemented products in pulp, paper, fertilizer, refinery, ethanol, methanol and chemicals productions. It can be deployed in nearly any process.
To learn more about ASSET360, download our solution brief.
This is a published conversation between:
- Darrell Proctor, Associate Editor at POWER Magazine
- Brady Kirkwood, Lean Monitoring & Diagnostics Engineer at EthosEnergy
- Dave Brill, Associate Vice President of Asset Management Services at Black & Veatch
- Paul McRoberts, President of AtonixDigital
- Webinar listeners