Our CTO, Tim Lieuwen, joined Tim Echols and KC Boyce in Energy Matters Podcast – WGAU to highlight on-going energy and sustainability research at Georgia Tech, and shared his summer experience in Alaska – a mountain bike, GPS receiver, and singing to bears.
You can listen to the full episode at Energy Matters – Ep.81
I can count three recent experiences where we get a phone call that goes something like this:
“We just did an outage and now we make less power than before, what’s going on?”
A couple of things can cause this to happen. Most gas turbines are controlled to a constant firing temperature, but the firing temperature isn’t measured. In reality, a combination of compressor and exhaust measurements are used to estimate the firing temperature. That estimate is typically what is shown on your control screen or saved to the historian. The other important thing is that estimate is only good for a certain machine with hardware of an existing quality. If refurbished hardware isn’t up to standards, instrumentation was re-calibrated or replaced, or even in some cases if improved hardware is installed the assumptions for that estimate aren’t 100% valid any more.
Thus, while your control screen may say that your firing temperature is the same as before the outage, that might not be the case.
Analysis tools and techniques can be used to identify areas to investigate to determine if the problem is control or hardware related. Then a course of action can be suggested to restore performance to expected or required levels.
A lot of our recent work has been handling of instrumentation accuracy issues and their impact on M&D decisions. When something looks off, first check the calculations. Then, if things still look “off” the next step is to check obvious things like units, and gauge vs. absolute pressures. Once you’re convinced the units are consistent, have you looked for data recording issues? Are their compression artifacts or signal dropouts? Finally, have the instruments been calibrated properly? Do they show signs of drift, low resolution, or improper readings in the case of redundant measurements?
In short, using advanced analytical tools opens up a wonderful world of knowledge, but the sensor inputs must be verified and checked.
In the midst of the coronavirus pandemic, working from home is the new reality for millions of US workers. Many companies are implementing voluntary or mandatory work-from-home policies, which means a lot of us are facing with an unusual challenge of working remotely for the first time and full-time.
Surely this era is different from working home for a day or two. Working from home due to coronavirus outbreak is unforeseen and could be extended for a certain period of time. You cannot easily go out and socialize in person, and are limited only to essential activities. According to Business Insider, some of the biggest challenges of remote workers include struggling with loneliness, managing their time, and communication among colleagues.
I do too, had a difficult time transitioning myself to work from home for a few weeks. Then, I came across with these tips to improve my WFH experience to stay productive and healthy while working remotely:
Dress like you are at work, or at least from your pajama, helps even if you won’t be interacting with another person all day. By dressing up, your brain knows that it’s work time, and prepare yourself to focus effectively.
It is very tempting to work from relaxed environment at home, but this could take a huge penalty on your productivity. With a designated separated workspace, you will be more alert, more confident, and more organized.
Just as you designate and separate your physical workspace, it is important to distinguish when you are working and when you are not. This is the important step to ensuring productivity while working from home. Setting yourself consistent hours keep you accountable and productive. Moreover, it is your best practice to keep your regular hours on the same schedule as your coworkers if your roll is collaborative.
Communication is especially important if you and your team have to go fully remote. Regularly communicate with your team when you convey any changes or new assignments to one another. If you encounter unique challenges as you try to do your job remotely, reach out to your team for your help.
Before you start working, a little habit of planning your work day will help your productivity up. Check what your priorities are for the day, how long you think it will take you to get everything done, and what you will work on if you have extra time.
For an example, here are the breakdowns of my new working-from-home routine:
8:00AM
Wake up, do some morning stretching, have a light breakfast, and dress up. About ten minutes before working, I jog down my plans for the day.
9:00AM – 12:00PM
Work, work, work! I have separated my workspace from any distraction in my house. For example, when I’m near my refrigerator, I tend to open and close the door repeatedly staying hungry. My home office is the farthest from the kitchen.
12:00PM – 1:00PM
Prepare and have a lunch, and take a short stroll around my apartment complex. I use this time to check my mail box.
1:00PM – 6:00PM
Work, work, work! Based on a short personal analysis, I am much focused between 2PM to 6PM. Every hour or so, I take few minutes to stretch my body and to look out the window so that I could rest my eyes for a bit away from the computer screen.
Following this routine helps me to stay more productive while I can keep practice healthy social distancing life.
So top tip, sometimes we tend to overlook the obvious when dealing with data. For example, Europeans use ‘,’ instead of a decimal. Really wrecks havoc when you’re importing files and don’t take that into account. Even when you’ve been doing something for 15 years don’t forget to take a step back when solving a problem.
This week I was working on a test recommendation report for a customer. This reminded me how important it is to have the required data available and ‘ready-to-go’ before undertaking any advanced analytics projects.
Understandably, most organizations want to jump into ‘big data’ without first realizing that the first step is to organize, catalog, and standardize data sets.
Machine learning and Artificial Intelligence is only as good as the underlying data it must learn from. Skipping this cumbersome, but necessary preparation is why many advanced analytics projects fail.
This is also why more than half of the report I was working on contained recommendations for data fields to store, storage frequency, naming conventions, and a common database for joining related, but disparate information sets.
© 2024 Turbine Logic. All Rights Reserved.
