Mark V-e or Mark VI

I am sure there will be somebody in another part of the world thinking of the same thing. It will be very hard to decide which one to go with with the limited information available especially if the info mainly comes only from GE.

Mark V is reaching its end of lifecycle very soon and there are 2 main choices whether to upgrade to Mark V-e or Mark VI. Our plant was commissioned in 2003 with Mark V as the Single Shaft Turbine controller. Considering the plant is still new, and since after implementing some modifications required by the past TILs and still there are more to come, we are running out of available memory / IO capacity even alarm tag allocation capacity & etc very quickly and we foresee that the upgrade will be imminent.

Appreciate if somebody can share your opinion which is the better upgrade option.

Thanks.

It certainly sounds like the Mark Ve is a good option for you; you get the same processing and I/O expansion capabilities as the Mark VIe without having to change your existing terminal boards or cabinets. Basically the Mark Ve upgrades the processor cores to the same hardware and runtime software as the Mark VIe. The HMI is also upgraded.

TurboYankee posted about a Mark Ve upgrade last month: http://www.control.com/thread/1026247032#1026247823

We have four 7FAs, one of which has been running a Mark Ve for the past year or so. I don't have the statistics on all of the Mark Ve problems, but I just returned home from a another callout on the Mark Ve. In fact, my last callout to the plant two weekends ago was for the Mark Ve.

Tonight's issue was the <C> core failure. The unit was cruising along at full load when the HMI alarm screen was deluged and the CT walked itself offline and parked at FSNL. The Ve would not respond to any commands, and the displayed data did not match the real world (e.g., the HMI showed fans off that were running, etc.) So it was obviously a <C> core problem.

I had to cycle power on the <C> core at which time the CT flamed out and ended up on turning gear. At least it wasn't a full load trip this time, as it has usually done in the past.

Two weekends ago the <T> core failed during operation. The CT did not trip, and marketing shut us down late that night so I was able to power down, <T>, then <S>, then <R>, the three <P> cores one at a time, followed by <C> before the <T> core alarm failure would clear. This, of course, had to be done off line as each time I power cycled a core it seemed like every contactor in the PECC dropped out and picked back up again.

Our reason for purchasing the Mark Ve, one of the first and only in the US, was that site management decided to purchase the casing management system which requires the higher capacity of the Ve. Casing managment cools the casing to reduce blade tip clearances which provides increased efficiency, so they say.

The Mark Ve has been a welcome source of OT which I am using to supplement the company contribution to my engineering degree. So, from a selfish perspective, I'm rather fond of the Ve. But would I upgrade the other units if I was higher up the food chain? Not yet. Though there's been improvements in Ve reliability, there's still too many bugs to be worked out.

The fact is we don't have near the problems with the Mark V's that we have with the Ve and we hardly use the casing management system because our load swings too often. (Casing management is only enabled when load is relatively stable.) So, we've gained little at the expense of a lot of trips and run-backs.

On the positive side, the Ve gas valve and IGV LVDT calibration routine is easier and has a nifty trend screen that displays while you're doing them. Apparently it's just like the Mark VI software package. Downloading the Ve is also much easier then the V.

Probably the main advantage to the Ve upgrade is lower cost and turn around time vs. upgrading to the VI. And if you want to buy the $1,000,000 a copy continuous tuning package you'll need the Ve for that, too, or the VI...

I found the perspective of my plant manager's (PM) reaction to last night's runback on the Mark Ve to be an interesting contrast to my own. He wasn't bothered by the runback because, in the big picture, it didn't amount to much of an availability loss and it only counted as one start/stop cycle. Whereas I see it as an issue of lower overall reliability of the Ve when comparing it to the Mark V's.

In fact my PM is more worried about GE's plans to fix the problems than he is about the problems themselves because there is always risks when you tinker with these sorts of things.

Really, there are three main choices: Mark Ve, Mark VI, and Mark VIe.

In my personal opinion the decision depends on the answer to the question: How is your plant operated?


The biggest benefit of the Mark Ve conversion is the short outage duration time involved in converting the panel to a Mark VIe-compatible system. Because no field wires are de-terminated, and the Mark V isn't removed and replaced with a new control system, and no field wires need be re-terminated, the outage duration is very short. The Mark Ve uses the Ethernet-based UDH for communication between the turbine control panel and the HMIs, so a new communications network infrastructure must be installed to support that, but that's usually just a couple of new network switches and some Ethernet cable (UTP and possibly fiber optic).

The Mark Ve conversion involves replacing a few of the cards in each processor (<C>, <R>, <S>, and <T>) and converting the internal- and external communications from ARCnet-based to Ethernet-based. Is replacing a handful of cards and leaving the majority of Mark V cards in service (including the TCQCs, the cards in <P>, and the cards in the digital I/O cores) worth the short term gain of reducing outage time, which is what one of the biggest benefits of the Mark Ve seems to be? Sure, the Mark Ve conversion provides more memory, faster processors, and expandability (but that will mean a new control panel enclosure has to be mounted somewhere for more I/O capability). But, there's also all the existing Mark V cards which aren't replaced during a Mark Ve conversion.

Does GE have a solid plan for upgrading those cards now or in the future? Have you actually counted the number of cards that will be replaced with the Mark Ve conversion versus the total number of cards in the panel? Have you considered which cards will be replaced versus which ones won't be replaced? Consider the functionality of the cards which aren't being replaced, and ask about firm commitments to provide new "Mark Ve" replacements before the Mark V cards are obsoleted (which is still likely more than ten years away, and even then replacements will still likely be available from some sources, just like Mark II and Mark IV cards are now).

Can the plant operating plan/philosophy handle the two to four weeks it would take to remove the Mark V and install a Mark VI (or Mark VIe, since if you're going to upgrade, why wouldn't you upgrade to the absolute newest product with the longest life expectancy which would be the Mark VIe at this time)? This would involve loop-checks and re-commissioning after removal of the Mark V, installation of the Mark VI (which is physically *much* larger than the Mark V), and re-termination of all field wiring. While physically smaller than the Mark VI, the Mark VIe would still require de-terminating and removing the Mark V, installation of the Mark VIe, re-termination, loop-checking and commissioning. But, it's all of the newest hardware and software, including the HMIs.

If the answer to the question of whether or not the plant can be shut down for two to four weeks is no, the plant can't be shut down for that length of time, then the choice is Mark Ve. If the answer is yes, the plant can be shut down for the removal of the old control system and installation, loop-checking, and commissioning of a new control system, then the choice should be Mark VIe.

If it were my money and it was absolutely necessary to upgrade to achieve some desired functionality or to implement some absolutely necessary TIL, I'd go for the Mark VIe. Even if it meant paying some extra money to ensure a well thought-out conversion plan was developed to be executed in as short a time as possible. That could be helped along by identifying all the sensors in each of the various compartments and skids and locations and having teams of people get familiar with their location and method of loop-checking, which would greatly reduce the over-all time of conversion. Proper wire-marking during determination also helps tremendously during retermination.

Again, the changes for either the Mark Ve or the Mark VI or the Mark VIe would include new Ethernet systems for control panel and HMI communication. The displays would be similar to what you're probably accustomed to.

One last thing to consider in the decision to shut the plant down for the change to either Mark Ve or Mark VIe is that trying to do the Mark VIe conversion at the same time as a major inspection would likely be difficult if not a mistake. Just too many things going on at once, and the need to do loop-checks and commissioning at the end of the outage, when the outage may (will likely) run long and schedule considerations will begin to dictate "priorities" (be they right or wrong), will make for a very stressful situation and lots of poor decisions about commissioning.

But, wouldn't you agree: It depends on how the plant is operated and if it can endure the loss of generation (electrical and steam) that a full control panel conversion would entail?

In my experience, quite a few of the Mark V panels were installed very poorly. Proper signal level separation wasn't maintained when routing cables and wires, which led to lots of electrical noise problems (nuisance Diagnostic Alarms, etc.). Further, many Mark Vs weren't wired very well, leaving lots of wire coiled in the wire troughs behind the terminal boards and mixed and left unseparated in the bottom of the panels. Many of the installations weren't properly grounded, or an attempt was made to try to isolate the "instrument" ground from "protective" ground, which the Mark V was never designed to be capable of without some major modifications. Many installations have inexpensive 125 VDC battery chargers which aren't properly maintained and which are switched to equalize mode while the Mark V panel is powered, usually while the turbine is running. Some installations used AC-to-DC converters supplied by inexpensive inverters, and the converters do *not* provide any level of filtering to speak of.

Any control system change is going to have issues if signal level separation wasn't maintained during construction, or the grounds are not properly installed or maintained. It's impractical to re-route cables and conductors, so those issues will remain, and might even be made slightly worse with a more sensitive control system. Battery chargers can be replaced, as can AC-to-DC converters. If it's thought that a Mark Ve or a Mark VI or Mark VIe will solve or reduce any of those issues or problems, there are simply no guarantees of any degree of improvement or resolution, even if a different manufacturer's control system is installed.

So, please review *all* the reasons you think you might be wanting to use to justify a control system conversion or retrofit, consider how the plant needs to be operated, and consider what the benefits, primarily long-term, will be versus and short-term benefits. I know this sounds like I'm avoiding making a recommendation, but I did. I just want to mention as many of the reasons I've heard people use when considering to replace their Mark V panels, and try to point out what can reasonably be expected to be achieved. Because there are *many* reasons to perform a control system conversion, or retrofit, not just get the newest technology; however, only some of them can have realistic expectations.

These are just a few of the considerations from my perspective; I look forward to those of others.

Hi there,

Correct me if I'm wrong, since Mark Ve/VI/VIe have the same new controller/CPU, I'd presume your remark above applies to Mark VI/VIe as well.

P.S. Are you from Mountain View?

Okay; you're wrong. The Mark VI and the Mark VIe do *not* share the same controller. The Mark Ve and the newest versions of the Mark VIe share the same controller.

dear CSA

u seem to have lots of experiences with Mark V/Ve/VI/VIe..

we currently have 2 Mark V panels, 1 for GT + 1 for ST.

From my understanding, what Mark Ve does is actually a "brain transplant", i.e. you'll have the processing power of Mark VIe family.

Being a "half" Mark VIe product, Mark Ve can utilize one nice feature of Mark VIe - distributable I/O.

1. Does that mean we can "extend" our I/O (if need be) by having new I/O panel and then link it up to the main controller panel via Eth/Fiber?

2. Do u have any experience deploying OPC server for Mark Ve/VI/VIe for the purpose of 3rd party historian system ie. PI system?

3. In our case, Mark Ve come into picture due to several reason:

- ~~ 50% cheaper than Mark VIe (need to manage the cash flow)

- shorter outage time

- complexity of upgrade to Mark VIe. based on the outage window, the conversion would coincide with our Major Inspection for GT/ST/Gen. just can't imagine how tedious it would be.

- GE's positive commitment/plan to address obsolescence of remaining parts after conversion to Mark Ve. GE's comment on this issue are as follows:

------------------------------

Phase 2 is a new product in the definition phase to address the 2014 obsolescence of the Mark V I/O electronics. Mark V I/O is left undisturbed during the Mark V to Mark Ve upgrade. The intent is to replace the electronics, most likely the ribbon cables and potentially the terminal boards. Product development is just beginning, any input you can provide on what solution best fits you expected needs is welcome.

If space is an issue, the Mark Ve is the best immediate solution to increase computation power and enable expandability of I/O. Phase 2 is in development and will not be available until late 2010 or 2011. The sole purpose of Phase 2 is to address all remaining obsolescence issues the Mark Ve does not address in preparation of the Mark V 2014 obsolescence.

We are currently defining what exactly the phase 2 solution is. The question is: as a user, which solution best fit your needs.... (1) an upgrade of electronics only requiring up to 2 outage days to install and checkout, (2) an upgrade of electronics and ribbon cables requiring from 2 to 7 outage days to install and checkout, (3) an upgrade of the electronics, ribbon cables, and terminal boards requiring 2 to 3 weeks of outage time to install and checkout.

-----------------------------
Regards.

Just wanna insert my query in this thread. We've all been exposed with the benefits of Mark-Ve in terms of the costing, new processors, new workstations and the upgraded interface between the control system. I'm just curious on the possibility of the Mark-Ve's compatibility between the system with different GE Machines. We are running on a Single Shaft Frame 9FA combined cycle plant. I've heard some rumours that there is a possibility of our machine can't be fitted with the Mark-Ve system.

We are still trying to get a couple of feedbacks on this matter from other plants and also GE. Is there any written articles on this matter or even any experiences from you guys that can guide me on this issue?

That's not true at all. The Mark Ve product has very few limitations in HDGT and ST applications. There is no reason it can not be applied to a 1+1 combined cycle single shaft application. If you have separate gas and steam control panels then this is just a typical application.

bh-4,

Have you received any responses from GE to your queries about Mark Ve suitability for single-shaft STAG applications?

There are also other options available, like Siemens PLC based controls systems (TCS). This system is developed by a dutch company and has proven itself by performing even better than Mark VI for almost half the price of its upgrade. More interesting is that all system parts are standard PLC and therefore you will be completely independent of OEM.