Blog
Latest Squawk: Thinking for a Change at Emerson Exchange- Dusseldorf
By Noelle Weymer on 06.06.12
Mimic Simulation

After the first European Emerson Exchange in Dusseldorf, the MYNAH team had a 15+ hour travel day back home to St. Louis, MO. This gave us plenty of time to reflect on the previous 2 days at Exchange.

On the plane, Application Engineer, Edin Rakovic and I were talking about everything we were able to do and learn while at Exchange. He started to jot down some notes and the next thing you know, this squawker received a 3 page essay from him! After some internal squawking, I decided I would share this platform with my colleague to share his thoughts…

Enjoy!

Thinking for a Change at Emerson Exchange 2012
By: Edin Rakovic, Application Engineer

I pride myself on discovering new approaches on solving problems in the automation industry. From day 1 at Exchange, my quest to discover new approaches had begun. It all started with the opening session in which I learned the importance of being involved in product and marketing innovation. My role at MYNAH Technologies has allowed me to participate in product innovation, but what is marketing innovation? As the session concluded I was able to come up with my interpretation of marketing innovation; marketing innovation is making sure that our customers are fully aware of the vast capabilities of our product line. (Squawker’s note: Silly engineers- that’s my job!)

As the conference continued, I was able to absorb new ideas and concepts. One concept which stood out was “justification”. Why automate your plant? Is there another solution? As I could not come up with one,I expanded this idea to “Why not have an OTS for your automated plant?” As I attended other presentations, this stuck with me. During his presentation, John Dolenc said, "don't compare against no solution, instead compare against another solution". This lead me to compare a dynamic process model completed in Mimic versus an empirical model implemented using linear equations. What happens outside the equation boundaries, typically slightly outside of steady state behavior? The answer is simple, it doesn't work! The solution is to use a first principles model that provides dynamic feedback across the full range. My next thought was, “Why are there so many of these empirical models out there in so many OTS systems??” This empirical approach must have appealed to someone?! This brings me to another quote from yet another presentation at Exchange. Roel Mudler said, "a 20 year old code in a calc block is not maintainable". That’s only the code! Will the engineer that wrote it be there? How long before your OTS becomes unmaintainable??

Operator Training Systems are unique and each one has its place. However, at MYNAH, we strive to provide a fully dynamic operator training system based on a first principles, high fidelity, process model. Do you need high fidelity? Yes! In a plant that produces multiple products, the switchover from one product to the next can vary between advanced operators and junior operators. A dynamic simulator can close this gap. During his presentation, John Dolnec also mentioned that, "most operators struggle running a batch for the first time." He continued to say, "some SOP's allocate steps during which the operator is supposed to tune.” When I heard, "tune" in a SOP, I was shocked that this situation still exists! As an engineer, my thought is that there has to be a way to tune it once, store it and let the CPU be "consistent".

What about the cost of a system shutdown? I learned the cost of a shutdown is composed of the following pieces: replacement hardware, labor to install hardware, process consequences clean up, value of production loss, etc. Can you afford this cost? John mentions that based on his research, he was able to conclude that "manual operations are done constantly with human error". I took this as a challenge to find the pieces of the Mimic software bundle that can contribute to preventing these issues. The solution is made up of the following pieces:

1) Use a virtual platform (VMWare) to build a virtualized OTS platform.

2) Setup OPC communications between DeltaV and Mimic.

3) Prepare your DeltaV database for simulation, reassign modules and run simulate convert if necessary.

4) Build a Mimic database using our DeltaV FHX utility.

5) Build a first principles model and tie it into the auto generated IO.

6) Take snapshots of your system in various stages using Mimic Operator Training Manager.

7) Create several scenarios for operator training conditions.

8) Build an instructor graphic from which you can launch all of the snapshots and scenarios.

9) Setup a VMWare snapshot of the whole dynamic virtual simulator.

That’s it! The solution is simple and can be achieved in fewer than 10 steps. With the proposed solution, we can change the statistic John mentioned when he said, "the key is finding people who can do the work onsite". Speaking of working onsite, the MYNAH team was busy preparing our exhibit…

On the 2nd floor of the "Saal Maritim" room, booth #3 was the MYNAH exhibit. It took us approximately 15 minutes to start a full working, first principles model of a distillation tower using DeltaV simulate and our Mimic advanced modeling package. The VMWare snapshot brought the proper configurations and database values, while the Mimic snapshots provided us with a steady state distillation tower.

During the exhibiting hours, we met with numerous customers and it was great to speak to them all. However, one customer stood out the most. I was approached by a Croatian customer who speaks Serbo-Croatian and asked for a demo of Mimic. Since I left Bosnia in 1996, and my Serbo-Croatian is, let’s say, small, if not extinct! Nevertheless, we understood each other and I had achieved a personal milestone; I was able present a MYNAH product in my native language! (Squawker’s Note: I still think they were just shooting the breeze in Serbo-Croatian…I guess I’ll never know!)

The MYNAH booth during exhibition hours at Emerson Exchange-Dusseldorf

Next, I took the time to see what options exist for plants running control systems that are older than me. (Squawker’s Note: Edin is only 25!) As a result, I found there are numerous plants in this position and Emerson has multiple solutions to help them migrate to a newer version of DeltaV. During this presentation, I discovered that plants degrade significantly over time and as the equipment ages, operations loosen up and outages increase. As Roel mentioned during his presentation, the key is continuous optimization. He discussed how automation reduces system variability alongside of the loop control. As I was letting everything he was saying sink in, a slide appeared that looked very familiar it had a MYNAH logo on it! It was clear that the solution of eliminating legacy control is made up of phases. One of the phases mentioned takes the control logic out of the PLC (GE, Modicon, Allen Bradley) and transfers it to a DeltaV controller. This is not surprising since the DeltaV controller outperforms these 30 year old PLC processors in every test.

The architecture in Roel's presenation (shown below) describes replacing the PLC CPU while maintaining the IO.

1) First, a MYNAH PLC IO interface is used to capture the remote IO. The PIO translates the remote IO protocol to ModbusTCP, which is a protocol that is carried over Industrial Ethernet.

2) Second, a MYNAH Modbus TCP/IP Virtual IO Module is used to communicate to the PIO across the Industrial Ethernet network. The VIM acts as a gateway between the industrial Ethernet network and the DeltaV railbus.

3) Third, a DeltaV controller is used to capture the IO from the 4 emulated program able serial cards. In addition, the controller will execute the logic which was previously being executed by the PLC CPU.

It's that easy and is just 3 simple steps!

PLC IO EtherNet Connection

On Thursday afternoon, it was time for me to present “Industrial Ethernet Integration with DeltaV”. It was an honor to be provided with the opportunity to represent MYNAH Technologies at Emerson Exchange. The room was small but we packed it full of customers from all around the world. There were 50+ attendees. (Squawker’s Note: If you missed Edin’s presentation, click here to view an online version of the presentation)

Edin Rakovic presenting Industrial EtherNet Integration with DeltaV at Emerson Exchange-Dusseldorf

Overall, it was a great conference and experience that really got my wheels turning. Thanks to the Emerson team for this great opportunity to really push my thought process to start “thinking for a change”!

Final Squawker’s Note: Thank you Edin for this great synopsis of your experience at Exchange. Also, to echo what Edin said, thanks to all the Emerson folks for putting together a great first European Emerson Exchange! It was wonderfully done. See you all in Anaheim!