As my power plant is 50 years old we want to go for complete automation. So the deciding factor for boiler & turbine control automation is whether to go for PLC OR DCS.
I'm working on a very similar project. The client request was to implement a DCS. But the main units (Gas turbines, HRSG's, Steam Turbine) must be capable of operate alone, if the communication with the DCS is loss.
The days when there used to be a major difference in DCS and PLCs are coming to an end. DCS and PLCs now have capabilities quite similar to each other (sophisticated analogue control in PLCs, and logic control/interlocks at DCS).
Infact, sometimes its hard to tell if a "product" is a PLC, or a DCS? For example, the GE speedtronic control system is really a mix of analogue control and digital logic.
So you really should not be worrying too much about whether or not you go for a PLC or DCS. You need to think about other (more important) things. To my mind, these are
1- Is a product (for example DCS?) targeted for your specific application?
For instance, it might be specifically targeting the power generation sector, with hardware/software optimized for that application. For a power plant, redundancy in digital/analogue I/O might not be so very important (but very important if its a for a process industry dealing in dangerous chemicals). That same DCS will know where on the power plant redundancy is needed (think turbine speed pickups), and they will provide redundancy for speed pickup cards. I hope I was able to explain it clearly.
2- Life cycle
Modern control systems (DCS, PLC etc) have a notoriously short life cycle. Give a top notch control systems 10 years, and it will join the ranks of its obsolete brothers and sisters (RIP). So go for a product thats proven in the field yet in the "green" part of its lifecycle (still in active production and with active support from OEM).
3- Consolidation
Another important consideration is control system consolidation. You want to minimize the types of control systems and PLCs on your plant. This reduces your total cost of ownership in reduced inventory, spares training and maintainability costs. This also helps in data communication (point 4 below). Wouldn't it be nice if one control system controls the Gas turbine, the boiler, the steam turbine and the BOP. Imagine the ease of maintenance and reduced TOC.
4- Ability to integrate with satellite control systems.
If the product is able to interface and communicate (not talking about hardwired I/O, but communication such as MODBUS) with packaged control systems elsewhere on your plant, this will vastly increase operator visibility. Ideally, even the packaged control systems should all be of the same type. That way communication can be the easiest.
5- Ease of maintenance and visibility
This is something that will be important to very many maintenance engineers. You need to be able to search control logics easily, view it online, control constants should be easy to tune, I/O point information should be detailed and easily available - all preferably from the operator stations. The engineering software should be easy to use. This is where traditionally PLCs lose out. DCS software (operator and engineering) is more sophisticated and feature-rich. This makes troubleshooting easier.
Off the top of my head, I would consider all of the above and wouldn't worry about whether its a PLC or a DCS.
You can investigate the Moore Quadlog safety PLC for the boilers in combination with the Moore Apacs DCS and stand alone PLC's for the power turbines in conjunction with a small local SCADA just for local monitoring. For a overall control and monitoring of all the plants and utilities use the same Moore Apacs DCS.
Virtually impossible for these systems to fail if you have a UPS as backup. Reliability and versatility are very high on these systems. Dual redundancy on top of dual redundancy.
These are being used all around the world on offshore installations which are some of the most hazardous and critical operations around.
Look at other companies and see what they have done and chosen, it will help in your final decision what to go for. Above is only one example.
Basically reap the benefit of someone else's hard work and investigations.
Thanks for the information but that is most probably true of just about all the systems installed today. That's how it goes in this business. It was still brand new and top of the range this morning and that same afternoon nobody can supply me with spares anymore.
Anyway the point I wanted to get across was to look at what other major companies have decided on so there is no need to do the whole investigation yourself.
As many of the replies confirmed that the difference between PLC and DCS is diminishing gradually.
As you mentioned that application being Boiler and Turbine control, please consider the below aspects while down selecting:
1. Controller scan period, very important as the process includes a rotating equipment at 3000 rev/s. So better the scan period, precise the control and interception.
2. Controller capability to Sequence of Events capturing at the lowest possible time (1 msec would be good)
3. System capability to support HART enabled devices - for the ease of PM.
4. System capability to support FIELDBUS (Profibus or Foundation Field Bus) - to reduce the initial capital and subsequent maintenance.
5. System capability to support next gen control solutions comprising of MBC (Model Based Control) and MPC (Model Predictive Control)
6. Diagnostics capability of the system hardware to the lowest level as possible (I mean, obtaining A/D conversion module diagnostics would be a great thing!!!!!)
You also need to consider some different technologies that can potentially reduce the cost of the upgrade. You did not state the existing control system types, but if your existing controls include pneumatics (quite likely in a 50 year old plant if there has been no previous upgrade) you should consider a Foundation Fieldbus based replacement for the pneumatics. Compare and contrast to a 4-20ma/Hart solution with remote IO, looking at the total installed cost including wiring and work that can be done on an extended routine basis before the outage starts.
We upgraded a 40 year old pneumatic boiler control system in 2006, using plant electricians who (months prior to the outage) ran a few home run armored fiber optic cables (FF HSE) from the controllers to four key plant locations. From small panels at those locations they pulled armored twisted shielded pair cables (FF-H1) to new FF transmitters located next to the existing pneumatic transmitters. All of the armored cables were tie wrapped to the existing building infrastructure, since there were very few cable raceways and conduits in this pneumatic based plant. The devices were verified for proper communication before the outage. Then during the outage the pneumatic transmitters were removed, and the new FF based transmitters were mounted while the electrical connection was already in place.
If you decide that a Foundation Fieldbus based solution is the best for the upgrade to the pneumatic controls portion of your overall scope of work, that needs to be a consideration when looking at your control system provider. I don't know of any PLCs that are in that space, but I might be wrong. Not that hybrids are a problem (there are pros and cons to every decision, no?), but that in itself is another key question. Do you want a single solution or a hybrid? If a hybrid, what is the risk and who is responsible for mitigating it?
Thus far we have two basic questions (technology and one vs. multiple suppliers).
An additional basic question is in regards to experience of the supplier(s) to your type of equipment. Boiler controls and burner management systems are pretty routine. But if you are, for example, considering upgrading an MHC controlled turbine to EHC controls, with the associated hydraulic power unit and other goodies, and you want to use the same controls supplier for the servo valve/LVDT interfacing, this again might limit the vendor playing field.
Yet another consideration is the experience of the supplier(s) to your specific applications and in your neck of the world. In this category I would also add that you should consider the resources for on site technical support.
After these and issues that others have/will post are considered, and you are comfortable with the supplier(s), who cares if it is a PLC or a DCS? I wouldn't.
Your post about upgrading a 40yr old pneumatically controlled plant is EXACTLY what I am trying to accomplish right now. We have two steam plants built around 1973 with mostly Bailey pneumatic controls. I'm new to the utility (been there 1.5 month), but they have had some vendors come in and give some quotes in the past few years. However, nobody in the utility really had much "plant experience" or control experience.
I'm currently reviewing all these proposals. High-ups balked at the price, but they don't understand the benefits. Additionally, prices have come down the last year or two due to the economy.
I'm not a control integrator by any means, but through various jobs I've picked up DCS knowledge. I'm most familiar with Ovation.
My main dilemma is this: The turbine control was upgraded to GE MkV around 2001-2002 with EX2000 excitation. It's an EHC turbine. The system works well and I plan to keep it running for years to come--the only upgrade would be 3rd party HMI's (see other thread).
However, a full boiler control upgrade may be in the future. In my opinion, it is best to integrate all Turbine, Boiler, Burner, and BOP systems into one DCS for ease of operation and ease of maintenance. In a situation like this, would one continue using the MkV turbine controls and communicate with the DCS or use the DCS with new turbine controls?
The MkV works fine and we have plenty of spare I/O and processor modules, but it is "obsolete". Budget's are always a concern, but I don't want to put a new engine in a car that has a frame that is about to rust apart....
> In my opinion, it is best to integrate all Turbine, Boiler, Burner, and BOP systems into one DCS for ease of operation and ease of maintenance. <
I absolutely agree. I've had the pleasure of seeing well over 10 original equipment turbine control systems hit the dumpster over the last 15 years. Each time we integrated with one DCS and it has been great. There is more than enough experience outside of the turbine OEM to do the control systems in a DCS.
> In a situation like this, would one continue using the MkV turbine controls and communicate with the DCS or use the DCS with new turbine controls? <
Out with the old. Sell on Ebay. By the time you get the DCS system approved, designed, and installed the obsolete stuff will be that much older.
> The MkV works fine and we have plenty of spare I/O and processor modules, but it is "obsolete". Budget's are always a concern, but I don't want to put a new engine in a car that has a frame that is about to rust apart.<
In some cases we had more money in OEM spare parts than the new DCS turbine controls cost! I would assume like us you are paying tax on that inventory. Look at all of the angles when trying to justify the project.
The only gray area from my perspective is the ESX. We have about 9 and have not replaced one yet. Ours are on combustion turbines and we have done some HMI upgrades. The aforementioned OEM turbine controls upgrades were on steam driven turbines (100 GMW to 850 GMW range), none of which had the ESX.
Ours is a case with around 10 year old power plant. There are separate PLC for boiler and turbine controls (Allen Bradley PLC5/40). The turbine control is through WOODWORD make governor control unit. All PID loops (there are 10 in total) are controlled using YOKOGAWA YS-80 series controllers.
Rockwell people have approached already with RS-View32 for creating a HMI that would take care of integrating everything except, WOODWORD GOVERNOR, can somebody assist me exploring alternatives available?
A snap acting pressure switch, or indeed any simple switch with no other circuitry involved, is considered as "simple apparatus" as it cannot store or generate electricity. It does not need to be certified for use in an IS circuit. However, it does need to be protected to avoid excessive voltage across the contacts when open, or excessive current through them when shut, either of which could result in a spark capable of causing ignition. So a barrier is required in the circuit.
Regarding your second question, all instruments or components connected to the circuit downstream of the barrier (ie, on the side connected to equipment in the flammable atmosphere) must be certified for use as Intrinsically Safe - even if it is actually located in a safe atmosphere. This is because any such device can potentially store electrical energy, then discharge it into a fault in the flammable atmosphere.
Facebook
Google
GitHub
Linkedin
