Interfacing with a GE Mark VI

J

Thread Starter

john

We are discussing design specifications for a combined cycle plant, and want to control the operator environment, including but not limited to a "high performance" HMI.

Assuming that GE still delivers their HMI like Henry Ford delivered the model T ("you can have any color as long as it's black"), the goal is to put an HMI that we can control on top of everything in the facility, including the Mark VI should GE be part of plant (decision not yet made.)

I would like to avoid the classic interface to the GE HMI. Interfacing directly to the control system, like the GE HMI does, is preferred. Redundancy is also preferred. Any suggestions on interfacing?

Also I heard last week that GE charges for "every alarm" that is subscribed for. I would think that the proper term would be "every tag" and not simply just for alarms. Is there any truth to this?

Finally, any other pitfalls to avoid?

Thanks, John
 
Everyone wants the same thing. No one wants to use the OEM's HMI, preferring instead to use one of their own preference. Since there is no "standard" for commands and alarms/events it would be extremely difficult for GE to document their protocol for commands and alarm/events such that any HMI software could be adapted to work with their protocol.

Imagine you are supplying a multi-million USD piece of equipment and someone configures a command improperly and the equipment is damaged or destroyed because one particular scenario wasn't anticipated or tested when commissioning the plant (I know; these things NEVER happen, but let's just say it did for the purposes of this analysis). Who's at fault? If the Customer insisted on this HMI and didn't insist on full and complete testing (which costs more than using the OEM's HMI) and didn't participate in the testing and commissioning by writing a proper specification to ensure full and complete testing of every possible scenario (which would also cost both parties much more for planning and testing), who's at fault? There are huge monetary consequences for all of these decisions, and they need to all be considered when making them.

Further, let's say there are problems with operating the unit or with the information appearing on the display of the Customer-demanded HMI. The OEM field service person travels to site to assist with the issue and is unfamiliar with HMI and HMI software and is unable to quickly resolve and troubleshoot the problem. Who's at fault? Who has the responsibility for solving this problem in a timely manner?

I presume you are referring to a Mark VIe, and not a Mark VI, as there are very few new Mark VIs being shipped these days (mostly only older units which were in storage for a couple of years or so or were special-order units built for special prices or because of some contractual obligation).

If you are talking about a Mark VIe, I'm told there is a MODBUS (serial or Ethernet is possible) interface where the Customer is allowed to send commands and obtain data (for display and archival) via a MODBUS connection directly to the Mark VIe--bypassing the HMI. This is about the only option, but it doesn't help with alarms/events.

Matrikon, whom I'm sure will chime in on this thread, has a partial Mark VI OPC interface (I'm told they have yet to solve the alarm/event riddle).

In any case, you are going to need an 'engineering workstation' with ToolboxST for troubleshooting and maintenance activities of the Mark VIe.

You could use a Mark VIe for your DCS and then you would have a common control platform for the entire plant, simplifying a lot of issues. Spares; training; troubleshooting; interface.

And GE does offer an OPC interface through their HMI and with the speeds of PC hardware these days you would use this to communicate to your HMI of choice, with the GE HMIs as a back-up. (I've been to several plants recently with this arrangement, and the operators are warming to the GE HMIs, in some cases choosing to use the GE HMIs rather than the other HMIs they had been using for years. Operators can be difficult to predict at times.)

Think long and hard about these decisions and their implications. You can have whatever you want, if you're willing to pay for it--and to accept the responsibility of your decision. I've been to too many sites to count over the course of nearly three decades where people have tried to deviate from the OEM's HMI offering and have had many, many, many issues which would not have occurred otherwise. You can have whatever you want--if you're willing to pay for it, both in the upfront costs of planning and testing, and in the downstream costs of troubleshooting and maintenance.

From what I've seen lately, the offerings from other major OEMs are pretty much the same when it comes to the interfaces to their turbine control systems. And the same logic above is likely driving their choices, as well.

I was recently at a plant with one of the most recent versions of GE's HMI offering, WorkstationST. I can report that GE is actually working at resolving many of the more common problems with their earlier HMI implementations. Adding new signal names/data to displays is much, MUCH easier than ever before. The WorkstationST Alarm viewer, though not well integrated into CIMPLICITY, has much more functionality than GE has ever offered before. On a multi-monitor HMI, the Alarm Viewer can be used very easily, offering lots of filtering options.

The new implementation I witnessed still uses CIMPLICITY, but the interface between CIMPLICITY and the Mark VIe is GREATLY improved. And anyone can modify CIMPLICITY displays or create their own displays and graphics as desired.

And, I'm told GE is now shipping a newer version of CIMPLICITY (Ver. 8.n, I believe) along with MS-Windows7 on new HMIs.

So, this is not your father's Model T.
 
CSA - thanks for the reply. I trust you are open for a candid response.

First off, we are many months away from pulling the trigger, so your recommendation to "think long and hard" is precisely why this question was posted here in the first place. Control.com is appreciated.

Also, as someone who has his fingers in a large fleet (just replaced a Mark 1, and have dozens of VIe's, and stuff in between); based on your reply, I think that you undervalue the benefit of corporate standards.

Yes, simply going with the OEM HMI is like pressing the Office Max "easy button". No risk, no extra cost.

But for those that deal with a large fleet, including plants that have no GE equipment - we see great benefit to a single solution.

"Everyone wants the same thing." Hello! I could not have said it better than myself.

If the OEMs, GE included, would ask "why" five times in a row, they might in fact get to the point where those of us reside. We are the ones who 5, 10, 15 years down the road need to keep this stuff working, and working well.

Moving to standards, it is technically correct to say that there is no "single" standard. But there options are not infinite either. Again, talking from an end user perspective, we tie systems together all of the time, using a variety of the various standards. However, the GE systems are, based on my experience and observation, hands down the most difficult. Why?

Whether my actions, in part or whole, play a role in the destruction a $500M boiler or a $500M turbine-generator is irrelevant.

I understand the risks involved, including warranty implications. That is why I'm limiting the inquiry to the HMI level. I'm not proposing an open DCS instead of a Mark VIe, for example. OK, inside my head I am, but I realize the uphill battle that would be.

My thoughts around HMI quality control would be that we commission the system on the DCS HMI, with an OEM HMI in the shadows, and with both DCS and turbine OEM field service engineers. Certainly everyone will be getting paid and act as professionals, correct? How often is the customer right - always?

Will this cost more up front? Absolutely. I would think that "everyone wants it" and "many are doing it" as ample proof that the need is real. How often is someone willing to pay more up front and take more short term risk for long term benefits?

As a "gear head" engineer, I often lament about "the management" that can't think past the end of their quarter. I categorically dump the decision to simply go with the OEM in a case such as this into that bucket. My nature compels me to inquire.

It is good to hear that the Mark VIe may be opening up a bit to modern ways, such as Modbus. Oops, Modbus came out in 1971. Strike that modern comment.

GE opening up their HMI to use OPC - now that is modern. OPC first published in ~1996 IIRC.

On a side note, regarding Matrikon, I too think that it is getting a bit old that every post that has the key words "Modbus", "Ethernet", "OPC", etc. seems to trigger a sales pitch.

Using the Mark VIe for the DCS would probably cut it if I were limited to one site. But that does not fit a fleet based user like me. When GE demonstrates a life cycle policy longer than 15 years and migration path that is not based on a rip and replace (R&R) mentality we can talk. If GE is interested in getting into this crazy DCS business, they are over 20 years too late. Many of us do have the common control platform you mention, with all of the benefits, in systems provided by companies other than GE. Hence our desire to use the HMI from our common platform on top of a Mark VIe system.

So in closing, I agree that if we chose to think outside of the box, we need to count the cost. But you would think that because so many want to and actually do so, GE might think about how to make it easier to do it. Wouldn't that be refreshing? Would it hurt if we actually looked forward to dealing with GE?

Before I forget, indeed we have moved from the Model T, through the 1971 (Modbus) Motor Trend car of the year.....Chevrolet Vega (ugh) and up to a 1996 (OPC) car of the year Dodge Caravan. John
 
John,

I've been on the commissioning side of this and I can tell you: It's not fun for anyone. There's so much finger-pointing and generally the field people (the ones it always falls on to make this "idea" work) don't have the knowledge of the internals and gut-level workings of the systems to make it go smoothly.

I believe GE still offers their GSM option (GE Drive Systems Standard Messaging protocol) which runs over Ethernet, but it does require an HMI in between the Speedtronic and the Customer's HMI.

Hopefully Mr. John Emery of GE who seems to monitor this site will see your post and add it to the Voice of the Customer input to GE.

Again, I'm not promoting GE. I'm just providing a field engineer's perspective from having suffered through too many of these "ideas"--which on the surface should be easy.

Are you saying that other OEMs are better than this than GE? Can you provide more details? Again, I think Mr. John Emery would appreciate details. This is the kind of thing he can point to to get new product- and technology funding to be able to compete better.

I can tell you from my Mark V experience with GE, where I went from being a novice PC user to getting all to familiar with BIOS settings and interrupts and batch files, trying to qualify every PC was a nightmare. "Plug and play" was a joke. Every Customer wanted to a PC that was supported by their IT department, but IDOS and TCI didn't run on every PC.

GE has a reputation as a conservative company with reliable equipment. Opening up the UDH (Unit Data Highway--which is how HMIs communicate with the Mark VIe) to any control system is very scary to them, and rightfully so. The HMI in between the Customer's DCS HMI and the Speedtronic is a firewall for them.

I'm not making excuses for them (they can do that well enough!), but, again just speaking from a field engineer's perspective having had to work with DCS vendors to make this seemingly easy idea work.

I don't disagree that GE HMIs are needlessly complicated and poorly documented (anyone who reads my posts will know that). And maintaining them is certainly no easy task.

Having done many "rip-and-replace" retrofits, I can say with some authority that the whole system needs rethinking. Designing a new control system without regard for how it can be easily retrofitted needs to change. But that's not just limited to GE; most of the OEMs have a similar problem when it comes to retrofitting their control systems.

I empathize with your--and others'--needs and situations. But, the reality is the implementation is not as simple as it seems. I've been on sites with other major OEMs and they seem to struggle with the same issues when it comes to trying to be controlled and monitored via a third-party DCS HMI.

Again, hopefully your pleas and "suggestions" will not fall on deaf ears.

And, it's Staples' "Easy Button."
 
P
John,

I have just seen your post - we have a GE Mk VI interface to our
HMI/Historian. If you are interested, please contact me

.....Peter

--
Peter Clout
Vista Control Systems, Inc.
2101 Trinity Drive, Suite Q
Los Alamos, NM 87544-4103

(505) 662-2484
FAX (505) 662-3956
Cell (505) 450-7810

[email protected]
http://www.vista-control.com
 
Peter Clout,

It would seem you have some experience with GE EGD (Ethernet Global Database) communications based on your mill experience.

I think the question we are all interested in knowing the answer to with your GE Mark VI HMI/Historian: Can it display alarms and events with the controller time-tag?

We have some OPC vendors who can't seem to do that (alarm/event with time-tag) very well. That seems to be the part of the HMI that causes the most difficulty.

I've had a look of the website, and I'm curious, but I need to know about the alarm/event with time-tag function. I didn't really see any part of the site that dealt more specifically with the Mark VI interface.

Thanks!
 
Peter Clout,

I did find one reference to a "scanner" application that seems to work via GE GSM which must be running on a GE Mark VI HMI. That seems to imply that Vista doesn't communicate directly with Mark VI control panels over the UDH using EGD.

Can you confirm how your system communicates (via an HMI running GSM, or directly to the Mark VI processors via the UDH and EGD), and if it's capable of handling alarms/events with controller timestamps?

Thanks!
 
If you want to integrate Mark VIe into a non-GE HMI, it really isn't very difficult. You'll need GE's WorkstationST running on one or more (depending on scale) PCs. People frequently equate WorkstationST with the HMI, but that isn't the case; the Workstation software is totally independent and does not depend on CIMPLICITY. In a system without CIMPLICITY, WorkstationST would provide OPC DA for data and OPC AE for alarms, plus a ton of other common protocols.

If you don't use CIMPLICITY, you'll give up some of the integration improvements GE has made, like being able to drag logic blocks like MOVs from ToolboxST to an HMI screen and have the correct graphic automatically inserted and connected to the right variables.

I've also never heard of GE licensing by number of alarms, or points in general. They are actually somewhat unique in that regard--most DCS systems sell by point count.
 
Demigrog,

Everything you said is true. But the originator of the post doesn't want to communicate with Mark VIe control panels <b>through</b> any HMI or intermediate PC. He, and many others, want to communicate <b><i>DIRECTLY</b></i> with the Speedtronic (Mark V, Mark VI, or Mark VIe).

That's what everyone is searching for: A way to communicate <b><i>directly</b></i> with Speedtronic control panels for display and commands and alarms and events. They don't want to have multiple HMIs running multiple different HMI applications in their plant. They want a common HMI application running on multiple PCs communicating <b><i>directly</b></i> with every control panel/system in the plant--the turbine control panel, the DCS, the water treatment PLC, the burner management system, the gas compressor PLC, etc.

However, the unseen cost of creating all of the displays and testing all of the interfaces/protocols is never fully factored into this desire. And, worse, the effort that's required during commissioning to make all of this work is never fully factored into this desire.

On the paper contract, it should be easy. The salespeople all agree. But, until there's a way to have a common protocol that everyone accepts and understands (and OPC is NOT that protocol; MODBUS wasn't/isn't, either) then this is going to be a very costly and frustrating experience for everyone but those who want it to happen, wrote it in the contract, agreed to it, and think they paid for this to happen.

I like to think of GE HMIs as separate from the Speedtronic control panels. I'm referring to the CIMPLICITY implementation on the PCs used for programming, troubleshooting, maintenance, and operation (the operation done via CIMPLICITY displays). No one can get away from having a PC running the necessary software for programming, maintaining and troubleshooting a Speedtronic control panel (be it Toolbox, or ToolboxST, or the Mark V configuration/programming tools). An "engineering workstation" is what that can be called. (That's not exactly true because CSE Engineering's ITC can totally replace a GE engineering workstation and operator interface for Mark Vs.)

But for operating the turbine and monitoring operation and sending commands and viewing alarms and events everyone wants to use a different application than CIMPLICITY. Instead of using CIMPLICITY for the entire plant, they want to use another application.

But no one wants to go through a PC, running an OPC server, to operate, monitor or send commands to a Speedtronic. They want to do it directly to/with the Speedtronic using some application other than CIMPLICITY.

(GE Energy should do a better job of selling CIMPLICITY as a total plant control package instead of "pushing" it as the turbine control operating control package.)
 
B
We are installing a MarkVIe and will use EPM DCS to interface, building our own graphics and control sheets driven by combination of hard I/O and data linked points. GE Training has been slow to repsond. Do you know of a third party that can train ITs at a level to perform mapping from DCS (Ovation) to GE (Cimplicity)?
 
I was aware that the Mark VIe could be equipped with the ability to communicate via MODBUS without the need to interface through a GE HMI for data values and commands, but alarms and events were not yet possible via MODBUS. Has that changed? The problem with MODBUS is that there are no time stamps....

Or, will you be interfacing with the Mark VIe through a GE HMI running WorkstationST in order to get alarms (Process & Diagnostic) and data and send commands? I believe WorkstationST converts data, commands, alarms and events to OPC format for other devices to access/use. Time stamping, though is not usually done at the actual time, but rather at the time the data is made available via the.Mark VIe to WorkstationST.

Perhaps if you could be more specific about the interface we can help you identify a resource to help.

Also, best of luck with your efforts. I have to several sites which have tried this, and all have had high expectations but eventually "settled" for much less than they thought they were going to get. Much to the lawyers' enrichment.
 
Bob,

most power plant site I have seen that are using GE gas turbines and an Ovation DCS are using the GE "GSM" data to communicate. Ovation controllers can talk directly to any GE HMI configured to run GSM data. I have not personally setup a MKVI "E", but have worked with several sites using standard MKVI controllers, GE HMI's running Cimplicity 5.5 and 6.1. The newer toolbox ST sites use Cimplicity 8.2? I think.

But I think that the HMI can still be setup to talk GSM and should be able to talk directly to the Ovation system without any "mapping".
The other option would be to talk to the HMI or directly to one of the MKVIE controllers using modbus TCP. I have seen both of these used with the "E" system.

There are several companies out there that could assist you if needed. I can provide some suggestions if you like.
 
S

Svitlana Palona

John,

If you are still looking for a product that interfaces with GE Mark VI, please contact CSE Engineering (CSE) at [email protected] or call us at 925-686-6733.

Our <IBECS>/<ITC> solution directly interfaces with Mk IV, V, VI and numerous controllers, including Allen Bradley, Foxboro, Delta V, Ovation, Modicon and others. Our system offers data server redundancy, access to control constants outside of Mk VI programming environment,can act as an OPC server and offers historical data logging. More information can be found at www.cse-eng-inc.com

Please contact us and we will be happy to offer you a control solution that will fit your needs.
 
Top