CSA,

Thanks for being patient regarding this thread and our questions, I appreciate your persistence to help others.

And sorry for such a simple question (not dumb--as there are no dumb questions, right?).

While waiting for your response, I reconsidered my question and the conditions of the contacts in the rungs, in order to keep the machine running or trip it. I realized that it is related to "inversion masking" that you explained in first part of this thread.

I think it is quite natural for the one devoting his energy and time to explain things to others to expect that mistakes that has been pointed out and clarified should not be repeated. But Neo seems to be busy this days and I can't add anything to the stuff regarding cards as I'm from Mechanical background.

And regarding the other thread, I'm following it but I haven't read it more than once for now, I'll surely read the thread again and respond. By the way I've read this thread http://www.control.com/thread/1026245600#1026245640 and it is very helpful.

You helped a LOT in developing my understanding of GE-design heavy duty Gas Turbine control system. Thank you for making things simpler for us.

 

CSA:

Always glad to receive your response.<pre>
<b>L86GT</b>
l86tgt1 L86TGT1Z L86GT
----|/|--------( )------( )</pre>
l86tgt1 = Generator Differential Trip Chan 1
L86TGT1Z = inverted signal for l86tgt1

There is no rung for L86TGT1Z,and L86TGT1Z is just the inverted signal for l86tgt1. And i think information below may help:
J3:IS200TREG 1D5-Contact3-G1\l86tgt1

So it seems that l86tgt1 signal comes from hardware, and it is inverted in toolbox to meet GE's philosophy. I think we need your comment about this, and the suffix Z in the L86TGT1Z signal name is confusing.

CSA, can i ask one basic question:What exactly is the <b>Lock Relay</b>?
And where can i find <b>generator protection elementaries (schematic drawings)</b> to know all the things that can trip 86G-1?
----------------------------------------------------------------------------------------------------------
<b>L5ESTOP1_FZA</b>
<pre> L5ESTOP1_FBZ L5ESTOP1_FZA
----|/|--------------( )------</pre>

So when L5ESTOP1_FBZ is logic <b>0</b>, then L5ESTOP1_FZA is logic 1. I think if L5ESTOP1_FZA is logic 1, it means the emergency pb is pushed down.

CSA, L5ESTOP1_FZA is only used in the L4PSTX1 rung.
I checked E-Stop P/B in PPRO Backp Turbine Protection Module-TREG Turbine Emergency Trip,and do find it is a NC contact (I need to learn how to read circuits).

I hope that later we can go over the TREG and TRPG circuits as they are VERY important parts of the turbine trip circuit, which are hardware portions as opposed to software we are talking about now (application code software that runs in the control processors).

When we finish L4T
-----------------------------------------------------------------------------------------------------------
<pre>
L63QTX L45FTX L5ESTOP1_FZA L86GT L4PSTX1
------|/|------|/|------|/|-------|/|----------( )
</pre>
When lube oil pressure is too low to protect turbo machinery (L63QT=1), or there is an fire indication in (L45FTX), or the VPRO E-STOP PB is depressed (L5ESTOP1_FZA=1), or the current in three phases is different, then L4PSTX1 goes to logic 0, L4T goes to logic 1 ,L4 goes to logic 0, AND the GT trips.

Thanks to SB, he points out that signals in L4PSTXn rungs are NC contacts, that is the signal is logic 1 when it trips the GT.
CSA, i have seen the thread,and i respect your efforts.

As for Start-up Fuel Flow Trip, i am just wondered the causes and consequences of it.
---------------------------------------------------------------------------------------------------------------
I hope it is OK.

Best regards!
Neo

 

SB,

Some of the best controls TAs (Technical Advisors) I've ever worked with started as mechanical TAs. The control system isn't just the Mark [whatever]--it includes valves (control and manual!), and solenoids, and pumps and fans and fuel nozzles and spark plugs. And, before one can be a good controls technician one really needs to be a good operator--to know what's supposed to happen when, and how that happens (systems and components, included).

So, stop putting yourself down because you're not a "controls engineer" or that you don't have a controls background. There's not a university (except for possibly one in the UK) where one can really study and become a gas turbine controls engineer. That happens with OJT--On-The-Job training and experience. GE does offer some internal training courses for their design engineers, but it's pretty specialized.

Continue trying to learn controls and operation, and eventually you will reach your goal. Hopefully you're leaning a lot here on control.com. You're more than welcome to open your own threads to ask questions or get information.

 

Neo,

I may have completely missed something here....

You have twice typed (and I have once copied and pasted) the following?<pre>
><b>L86GT</b>
> l86tgt1 L86TGT1Z L86GT
>----|/|--------( )------( )</pre>
This is an "illegal" rung if indeed the element for L86TG1Z is a coil (as drawn) and not a normally open contact.

It's also a serious problem if it's <b>NOT</b> a coil but it IS a normally open contact from a non-existent rung. Let's not confuse inverted discrete inputs with "inverted" logic signals; they are not the same and it can be very misleading.

As was said in a prior post, a rung with a suffix of "Z" is <i><b>usually</i></b> a timer, as in the example we used of L4Z. L4Z was a time delay to energize--L4Z would be a logic "1" one second after L4 was a logic "1". I'm <i>presuming</i> that L86TG1Z is a timer, and that it will be a logic "1" some time after L86TG1 is a logic "1"--<b>but that's just a presumption.</b> It would take a rung with a timer block to write to L86TG1Z--it can't be done with the inversion mask of a discrete input.

If there's no rung or coil for L86TG1Z then there is a real problem. As we have noted, signals with lower-case alphabetic characters are <b>intended</b> to denote inputs and outputs of the Mark VIe turbine control system--and l86tg1 would appear to be that: a discrete input from a lock-out relay on the generator control panel.

A lock-out relay is a panel-mounted (or door-mounted) device with a "bat" handle, usually (it looks like a small pistol grip; I don't know why it's called a "bat" handle except that it may have looked like a bat wing in some incarnation in the past). There is a spring and a solenoid-operated latch in the device, and usually a long string of contacts (normally open and normally closed) on the back of the relay on a shaft to which the handle is also attached. When the bat handle is in the vertical position the relay is in the "RESET" position, and the spring is compressed and latched into a position that closes the normally open contacts and opens the normally closed contacts. When a signal from any of several different generator protective relays actuates, the solenoid of the lock-out relay energizes and releases the latch, and the lock-out relay spring rotates the bat handle and opens the normally open contacts of the lock-out relay and closes the normally closed contacts of the lock-out relay. The bat handle rotates slightly (usually in the anti-clockwise direction, or to the right, and the handle is no longer vertical--which is one visual indication that the lock-out relay has been actuated--that it has bee "tripped"). Sometimes there is an orange "target" which is uncovered on the lockout relay cover behind the handle which indicates the lock-out relay is in the TRIPPED position. Sometimes there are green and red indicating lamps on or above the lock-out relay cover, and the green indicating lamp is lit when the relay is in the RESET position, and the red indicating lamp is lit when the lockout relay is in the TRIPPED position.

There are a couple of purposes for using a lock-out relay. First, it requires a human to reset it--that prevents re-starting the unit until someone has resolved the problem and manually reset the relay. Second, one contact from one relay can be used to change the state of several contacts on the lock-out relay, which can be used to open the generator breaker, trip the turbine, provide alarm annunciation, and block other operations (by opening normally opened contacts in other strings).

Lock-out relays are a VERY important part of the protection of the unit--turbine and generator. Some generator lock-out relays are used to trip the turbine AND the generator; some lock-out relays are used to trip only the generator breaker, or some other device--but not the turbine. It all depends on how the generator protection is set up. Usually, the "main" lockout relay for the generator is called 86G-1, and sometimes it drives the discrete input l86g1. I've also seen it called 86TG-1, driving l86tg1 (with Belfort, we never know what they're going to do or name signals and devices).

So, let's just say the main generator lockout relay is called 86TGT-1, and it drives l86tgt1. A normally open contact of 86TGT-1 would be wired to the discrete input l86tgt1. The normally open contact would be closed when the lock-out relay was in the RESET position, and typically, l86tg1 is inverted in the IOPACK for the discrete input.

It would look like this in my version of "drawing":<pre>
INVERTED NO 86TGT-1 Discrete Input
to Digital Speedtronic Turbine Control Panel
(86TGT-1 Changes State (Opens) on a Differential Trip)

Hardware | Software
-------- | --------
|
Physical |
Contact |
86TGT-1 Input Screw l86tgt1
---| |----------o------------------(I)---
|
| l86tgt1
| ---|/|---</pre>
And here's the L86GT rung from Neo's machine, presuming the element is a normally open contact, not a coil:<pre>
l86tgt1 L86TG1Z L86GT
---|/|--------| |---------------( )---</pre>
If there truly is (and I believe you, Neo!) no rung writing to L86TG1Z then l86tgt1 can never trip the turbine! Because L86TG1Z will never close if there is no logic in a rung writing to L86TG1z.

Let's presume there was a rung for L86TG1Z, and let's further presume that it is indeed a timer, and let's say that it is driven by l86tgt1. It could look like this (I just put a value of 1.0 sec for the time; it could be any value):<pre>
l86tgt1 L86TG1Z
---| |--------------------------(T)---
1.0 sec</pre>
Now, let's analyze the L86GT rung presuming the existence of L86TG1Z as shown above. It would say that when there is NOT a Generator Differential Trip Chan 1 AND when there WAS a Generator Differential Trip Chan 1 for 1.0 sec then trip the turbine by making L86GT a logic "1". That's not "logical" (pun intended). If l86tgt1 goes to a logic "1" (when 86TG-1 opens), then the normally closed l86tgt1 in rung L86GT is going to open, and the normally open contact of l86tgt1 in L86TG1Z is going to close, and 1.0 seconds later L86TG1Z is going to go to a logic "1" and the normally open contact of L86TG1Z in rung L86GT is going to close. But nothing will happen because the normally closed contact of l86tgt1 is open--and the turbine will not trip.

Let's just presume, for the moment, that L86TG1Z is not a timer, and that is an "inverted" version of l86tgt1 (meaning that it's the opposite sense of l86tgt1). Then it would look like this:<pre>
l86tgt1 L86TG1Z
---|/|--------------------------( )---</pre>
If this was the case, then the turbine would be tripped all the time--presuming that the "coil" we've been typing and copying is a normally open contact:<pre>
l86tgt1 L86TG1Z L86GT
---|/|--------| |---------------( )---</pre>
because when l86tgt1 is NOT a logic "1" then l86TG1Z WILL BE a logic "1" which would close the normally open contact in rung L86GT. And, when l86tgt1 is NOT a logic "1" and when L86TG1Z IS a logic "1" (which it will be when l86tgt1 is NOT a logic "1"!) the L86GT WILL BE a logic "1" and the turbine will be tripped.

In my estimation the rung L86GT is seriously flawed. Presuming 86TGT-1 drives the inverted discrete input l86tgt1, if L86TG1Z is the inverted signal of l86tgt1 then L86GT would be a logic "1" all the time. And, I'm sure the turbine starts and runs; right, Neo?

So, the ONLY way the turbine can start and run is if there is no logic writing to L86TG1Z (regardless of what it's description is) so the normally open contact of L86TG1Z in rung L86GT would never close. The problem with that is that if 86TGT-1 ever opens because of a Generator Differential Trip Chan 1 and l86tgt1 goes to a logic "1" then the normally closed contact of l86tgt1 in rung L86GT will open, and because L86TG1Z can never close L86GT will never go to a logic "1" and the turbine will never be tripped by 86TGT-1/l86tgt1.

Either there's a typing error, or there are some serious errors in the application code for L86GT. There's no two ways about it.

Neo, is there a 86TGT-1 trip signal on the HMI Trip Display?

Is L86GT an alarm ("GENERATOR DIFFERENTIAL TRIP)?

Could L86TG1Z be a discrete input (with capital alphabetic characters instead of lower-case alphabetic characters)?

'Cause something's not right here. And this should be resolved, one way or the other. Either L86GT been "disabled" by using L86TG1Z which has no rung to drive the normally open contact in rung L86GT, or there's something very wrong.

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

I think we answered the question about lock-out relays above.

There should be a set of schematic drawings (GE calls them "elementary" drawings) for the Generator Protection Panel--the panel where the generator protective relays are located. It has to interface with many other panels and devices (PTs and CTs and breakers and such). These days, many of the older individual electro-mechanical generator protective relays have been replaced with a single digital generator protective relay, sometimes called a DGP. Sometimes, there is a redundant DGP, and sometimes there is another digital protective device that's used for step-up transformers. But, you should be able to find a set of schematic drawings (elementaries) for the panel which houses the generator protective relay(s). There are usually ammeters and voltmeters and synchronizing lights and switches on the GPP (Generator Protection Panel).

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

Well, we should go over the start-up excessive fuel flow trips; there's usually one for gas fuel, and one for liquid fuel. The purpose of them is to protect the turbine from high temperatures caused by an inability to control fuel during the firing attempt (high fuel flows during start-up). For liquid fuel, the liquid fuel flow divider feedback is used to check for excessive liquid fuel flow-rates. For gas fuel, the Stop-Ratio Valve (SRV) position is used to check for excessive gas fuel flow-rates. Yes; you read that right--SRV position. The assumption is that if the SRV has to go open past a certain point that the gas fuel flow will be excessive and so the turbine is tripped. That's right; gas fuel flow-rate feedback is not used for this trip--even though the alarm reads "Fuel flow-rate High". Again, the presumption is that the fuel flow-rate will be excessively high is the SRV is open past a certain point during firing.

But, we should verify the above by looking at the L2SFT logic for both liquid and gas fuel (I think you said your unit runs on both fuels, right?).

Let's get to the bottom of this lock-out trip issue, first, though, don't you think?

 

Neo,

I've been thinking about this, and this would be a valid rung--though it would not work (if l86tgt1 is inverted and driven by a normally open 86TGT-1 that's open when the lock-out relay is TRIPPED):<pre>
<b>L86GT</b>
l86tgt1 L86TGT1Z
----|/|--------( )
|
| L86GT
-----( )</pre>
If L86TG1Z is a coil, then it would be true that L86TG1Z was the "inverted" signal of l86tgt1. But, that would mean that L86GT would be a logic "1" when l86tgt1 was NOT a logic "1". And when L86GT is a logic "1", then L4T will be a logic "1"--which means the turbine will be tripped.

Let's go back to (the abbreviated) L4PSTX1:<pre>
> L63QTX L45FTX L5ESTOP1_FZA L86GT L4PSTX1
>------|/|------|/|------|/|-------|/|------( )</pre>
And L4PST:<pre>
L4PSTX1 L4PST
-----|/|---------------( )
|
L4PSTX2 |
-----|/|------
|
L4PSTX3 |
-----|/|------
|
L4PSTX4 |
-----|/|------</pre>
And here's L4T:<pre>
L4PST L4T
-----| |---------------( )
|
L4PRET |
-----| |------
|
L4POST |
-----| |------
|
L3SMT |
-----| |------
|
L4IGVT |
-----| |------
|
LFPAUXG2LLX |
-----| |------</pre>

 

CSA:

I bet you don't a good sleep last night!

1)
"I've been thinking about this, and this would be a valid rung--though it would not work. (if l86tgt1 is inverted and driven by a normally open 86TGT-1 that's open when the lock-out relay is TRIPPED):<pre>
<b>L86GT</b>
l86tgt1 L86TGT1Z
----|/|--------( )
|
| L86GT
-----( )</pre>
If L86TG1Z is a coil, then it would be true that L86TG1Z was the "inverted" signal of l86tgt1. But, that would mean that L86GT would be a logic "1" when l86tgt1 was NOT a logic "1". And when L86GT is a logic "1", then L4T will be a logic "1"--which means the turbine will be tripped."

CSA, i think you are right!

"Neo, is there a 86TGT-1 trip signal on the HMI Trip Display"
Yes, there is a L86TG trip signal on the HMI Trup Display.
-------------------------------------------------------------------------------------------------------------------------------------------
I checked these signals again in application code,and i think it would be better if i draw the rung as below:

<b>L86GT</b><pre>
l86tgt1 L86TGT1Z
----|/|--------( )
|
| L86GT
| |-------( )
|
| L86TGT1_ALM
| |-------( )</pre>
INVERTED NO 86TGT-1 Discrete Input to Digital Speedtronic Turbine Control Panel (86TGT-1 Changes State (Opens) on a Differential Trip)<pre>
Hardware | Software
-------- | --------
|
Physical |
Contact |
86TGT-1 Input Screw l86tgt1
---| |----------o------------------(I)-----
|
| l86tgt1 L86TGT1Z
| ---|/|----------( )
| |
| | L86GT
| |-------( )
|
| L86TGT1_ALM
| |-------( )</pre>
So,if l86tgt1 is inverted and driven by a normally open 86TGT-1, then <pre>
86TGT-1 NO l86tgt1 Coil l86tgt1 NC L86GT
Contact State Logic State Contact State State
------------- ----------- ------------- ----------
OPEN "1" OPEN "1" TRIP
LOSED "0" CLOSED "0"</pre>
But i can't find l86tgt1 is inverted or not in toolbox, nor do i make sure 86TGT-1 is NO contact.

Anyone who can help?
--------------------------------------------------------------------------------------------------------------------------------------------
2)
CSA, I read the thread you recommenced yesterday in a little hurry, but i find that l30sg1 is not inverted and 30SG-1 is an NO contact.

<b>30SG-1</b>
Diagnostic ignition exciter switch
characteristics: Normal
NO
125VDC

<b>l30sg1</b><pre>
SignalInvet Description
Normal Inversion makes signal true if Contact is open

l30sg1 l2tvx1 L30SG1_ALM
---|/|--------| |------(T)-----------( )
1.0sec </pre>
That is weird!
l30sg1 is inverted or not?
-----------------------------------------------------------------------------------------------------------------------------------------------------------
>And L4PST:<pre>
> L4PSTX1 L4PST
>-----|/|---------------( )
> |
> L4PSTX2 |
>-----|/|------
> |
> L4PSTX3 |
>-----|/|------
> |
> L4PSTX4 |
>-----|/|------</pre>
>And here's L4T:<pre>
> L4PST L4T
>-----| |---------------( )
> |
> L4PRET |
>-----| |------
> |
> L4POST |
>-----| |------
> |
> L3SMT |
>-----| |------
> |
> L4IGVT |
>-----| |------
> |
> LFPAUXG2LLX |
>-----| |------</pre>
<pre>
L39VT L2SFT L12H_FLT l4ct L45HA_T L45HT_T L4PSTX2
--------|/|------|/|--------|/|--------|/|--------|/|------|/|--------( )</pre>
Next time i will interpret L12H_FLT, l4ct, L45HA_T , L45HT_T, but to what extent?
As you said: Let's get to the bottom of this lock-out trip issue first, though,don't you think?

Are all the Discrete Input in L4PSTXn inverted? And all the related physical contact are NO contacts?
----------------------------------------------------------------------------------------------------------------------------------------------------------
That is a really long response of mine, and its not my style. I think you are changing me, CSA, and i must be careful.

And i will go on a short-term trip, during that time i may not have easy access to the web.

Best regards!
Neo

 

Neo,

I've been re-reading your previous post, and now I see that l86tgt1 comes from one of the discrete inputs to the TREG card--not a normal DI or DI/DO card:

J3:IS200TREG 1D5-Contact3-G1\l86tgt1

And this is where I'm making my mistake--the TREG discrete inputs can't be inverted. So, now everything works out--thought it's not a pretty sight.<pre>
l86tgt1 L86TGT1Z
----|/|--------( )
|
| L86GT
-----( )</pre>
OR:<pre>
l86tgt1 L86TGT1Z
----|/|--------( )

L86TGT1Z L86GT
----| |--------( )
|
| L86TGT1_ALM
-----( )</pre>
Yep; this works--but it's really ugly. REALLY UGLY.

In this case, a normally open 86TG-1 (that is closed when the lock-out relay is RESET, and open when the lock-out relay is TRIPPED) will drive a discrete input of the TREG card (which cannot be inverted), l86tgt1. So, l86tgt1 will be a logic "1" when there is <b>NOT</b> a lock-out relay trip, and will be a logic "0" when there <b>IS</b> a lock-out relay trip. Again, this violates the concept that logic signals will be a logic "1" when in alarm or trip condition. Which is why they are "inverting" l86tgt1 with L86TGT1Z.... L86TGT1Z will be a logic "0" when there is <b>NO</b> lock-out relay trip, and a logic "1" when there <b>IS</b> a lock-out relay trip.<pre>
Lock-out 86TGT-1 NO l86tgt1 Logic l86tgt1 NC L86TGT1Z L86GT
Relay Contact Coil Contact Logic Logic
Condition State State State Coil State Coil State
--------- ------------ ------------- ---------- ---------- ----------
RESET CLOSED "1" OPEN "0" "0"
TRIPPED OPEN "0" CLOSED "1" "1" - <b>TRIP!!!</b></pre>So, mystery is solved. It works, but it makes reading the application extremely difficult, and just violates about every philosophy and standard that's been in place for decades.

Well, we got to the TREG card a little earlier than I'd hoped, and it certainly wasn't very pretty, either. (There's still more to cover on the TRPG and TREG card--later.)

Mystery solved.

But I hope that people are beginning to understand that by following some simple conventions when writing application code and choosing signal names that reading and understanding application code--especially when troubleshooting--can be made much easier and simpler. And, when conventions are mixed in the same application code and configuration that it makes for some serious time-wasting exercises.

For my part, I apologize for not catching the TREG reference earlier and causing so much confusion. There are seven (7) "trip interlock" inputs to the TREG card available for chosen trip conditions to be able to trip the turbine via "hardware" (as we'll see later). These inputs were originally designed for SIMPLEX panels, but it seems the Belfort Bunch have unilaterally decided to use some of them in TMR applications, and do so with little forethought and poor description.

Anyway, lets' get over it--it is what it is, as a colleague used to say.

On to L4PSTX2!

Have a good trip. I, too, will be on another business trip in the next few days--with no Internet or cell phone access while on site (in a mancamp environment). I would like to be able to answer that "always [this]" or "always [that]" but, as we can see it doesn't happen like it should. We can generalize, and that's helpful--but one always has to be ready for the unusual--and I wasn't ready for this one; sorry!

Later!

 

CSA,

Thanks for the encouragement.

Yes, I'm learning a lot here. This is such a great platform. I’ve a good idea of field devices due to my Mechanical background and I'll continue to learn GE-design heavy duty gas turbine operation and control.

 

CSA


Can i ask one basic question before we go on L4T journey?

What is the definition of NO or NC of a physical contact?

We usually define it based on the state of the physical contact when there is no current flowing through the coil.

Best regards!
Neo

 

Neo,

Sure! You can ask questions--that's what we're here for.

Where to start? (Why is it that the simplest things can be the most difficult to explain--especially in writing?)

We can agree that a contact--whether it's a physical contact or a "virtual" contact in software--can be in one of two states, correct? That is, a contact can be either open or closed. And, that something (some force) is required to change the state of a contact. And that when that force is removed, the contact will revert back to it's "at-rest" or "normal" state. So, a contact has a "normal" or "at-rest" state, and an "actuated" or "not normal" state.

The strict definition of NO and NC refers to the state of contacts when a device (pressure switch; temperature switch; limit switch; electro-mechanical relay; etc.) in its box is taken off the shelf of a storage facility (warehouse; manufacturer or store shop) when the ambient temperature is 59 deg F (15 deg C), the ambient pressure is 14.7 psia (approximately 1 bara), and the relative humidity is 60%. In this case, there is no pressure (save for atmospheric pressure), no temperature (save for ambient temperature), no current, and no other force acting on the device. So, when that physical device is pulled off the shelf and taken out of its box at the stated conditions a normally open contact will read infinite resistance, and a normally closed contact will read zero resistance. The devices are said to be "de-actuated", and the state of the contacts in this condition (open or closed) is referred to as "normally open" or "normally closed", respectively. This is their "normal" state--their "de-actuated" or "de-energized" state. Their "energized" or "actuated" state is the opposite state (that is, closed for normally open contacts, and open for normally closed contacts).

Let's remember that a software coil is a virtual representation of an electro-mechanical relay (a physical device that changes state based on whether or not current is flowing in the coil of the relay), and that both software coils and electro-mechanical relays have contacts that change state when the coil or relay is energized or de-energized. So, when virtual current is flowing in a software coil it is said to be "energized". And, when current is flowing in an electro-mechanical relay it is said to be "energized" or "actuated". And, when no virtual current is flowing in a software coil it is said to be "de-energized" or "de-actuated". The same is true of an electro-mechanical relay--when no current is flowing in the coil of the electro-mechanical relay coil the relay is said to be "de-energized" or "de-actuated."

In the "de-energized" or "de-actuated" state normally open contacts are open--that is they WILL NOT conduct electricity (they have infinite resistance). In the "energized" or "actuated" state normally open contacts WILL conduct electricity (they have zero resistance).

In the "de-energized" or "de-actuated" state normally closed contacts are closed--that is they WILL conduct electricity (they have zero resistance). In the "energized" or "actuated" state normally closed contacts WILL NOT conduct electricity (they have infinite resistance).

This is true of BOTH physical devices and software coils/contacts. It's just that in the software representation, the current is "virtual", and in real physical devices it's actual electrons (amperes).

Does this help? Remember, a programmable automation controller is really nothing more than a modern substitute for physical wires and physical contacts and physical electro-mechanical relays. And, this virtual representation needs to--and does--have the characteristics as the physical devices it replaces.

If you have some specific questions about normally open and normally closed contacts of specific devices, let's answer them before we continue on this L4T journey.

 

CSA:
Thank you for your patient reply for NO/NC contact.
-----------------------------------------------------------------------------------------------------<pre>
L39VT L2SFT L12H_FLT l4ct L45HA_T L45HT_T L4PSTX2
--------|/|------|/|--------|/|--------|/|--------|/|------|/|----- ---( )</pre>
L39VT = Vibration trip
L2SFT = Startup Fuel Flow Excessive Trip
L12H_FLT = Loss of Protective HP Speed Inputs
l4ct = Customer trip input
L45HA_T = High High Gas level Det-Gas Fuel Moduel

<b>L12H_FLT</b><pre>
L14H_ZSPD L4 L14HT L12H_FLT
-----||-------||-----||----( )
L12H_FLT L86MR1 |
-----||-----|/|---------</pre>
L14H_ZSPD = VPRO Zero Speed Ind
L14HT = Cooldown slow roll start speed relay

So it means the speed signal is lost in VRRP card,and L4T will go logic 1.

<b>l4ct</b>
I have no idea of l4ct.
IS200TBCI 1D3 contact24

<b>L45HA_T</b><pre>
L45HA1_T L45HA_T
------||-------()
L45HA2_T |
------||----</pre>
L45HA1_T = High High Gas Level Det-Gas Compartment
L45HA2_T = High High Gas Level Det-Gas Compartment Vent Disch

I think when L45HA_T is logic 1,that means flammable gas is detected in DLN and other aux compartments, and for safety consideration is will trip the turbine.

<b>L45HT_T</b><pre>
L45HT1_T L45HT_T
------||-------()
L45HT2_T |
------||----</pre>
L45HT1_T = High High Gas Level GT Compartment
L45HT2_T = High High Gas Level Det-GT Compartment Vent Disch

I think L145HT_T is similar to L45HA_T.
-------------------------------------------------------------------------------------------------------
So, in conclusion, when the unit has a vibration problem, or the firing fuel supply is too much, or the speed signal in VPRO card is lost,or the gas fuel in DLN compartment and GT compartment is detected, then L4PSTX2 is logic 0,and then GT trips!

Is that OK?

It occurs to me that it has been more than three months since i post my first thread here asking how to get a better understanding of GT control, and you advice me to start with L4.

Thank you for your help, you get me interested in GT control, and i learned a lot from this journey.

Best regards!
Neo

 

Neo,

Hope the NO/NC explanation was clear. I can sort of understand how people can be confused by this, but it's really very simple. The problem seems to be visualizing that the NO or NC contact we see on a drawing or schematic or in application code seems "fixed"--and yet it changes state based on whether or not the coil or the device is energized or actuated.

And, I know a lot of people also get confused about "energized" and "de-energized" coils and contacts.

But, it's really pretty simple--and it's been around since the first days of electricity. And hasn't changed much. Sure, we have "virtual" coils and contacts--but that's just another way of doing the same things that real coils and contacts do. Just without so much wire and so many relays and physical contacts.

And, physical devices like pressure switches and temperature switches and limit switches and level switches and such also have actuated ("energized") and de-actuated ("de-energized") states.

Anyway, back to the L4 journey, and we're on the L4PST highway right now....

In a previous post you had written<pre><b>L2SFT</b>
L60FFLH L14HA L2WX L2SFTL
--------||-------|/|--------|/|--------(T)
2s</pre>
L60FFLH = Liquid Fuel Flow High
L14HA = HP accelerating speed signal
L2WX = Turbine Warmup Complete, Increase Fuel

The rung you posted was only the liquid fuel portion of L2SFT. We should also investigate the gas fuel portion of L2SFT.

<b>L12H_FLT</b><pre>
L14H_ZSPD L4 L14HT L12H_FLT
-----||-------||-----||----( )
L12H_FLT L86MR1 |
-----||-----|/|---------</pre>
L14H_ZSPD = VPRO Zero Speed Ind
L14HT = Cooldown slow roll start speed relay

This bit of logic compares the VPRO speed signals to the <Q> (<R>, <S> & <T>) speed signals, and generates a trip if there is a disagreement by more than one processor (<R>, <S> & <T> with respect to <X>, <Y> & <Z>--which are the names of the three independent VPRO cards in <P>).

Referring to L45HA_T and L45HT_T, I'm kind of surprised that the signal names are not in lower-case alpha characters--which would mean they are discrete inputs from a device or devices outside the Mark VIe. I know that GE sometimes uses Bently-Nevada or other hazardous gas sensors directly connected to the Mark VIe, and sometimes the sensors are connected to an external monitor which is connected via discrete contacts to the Mark VIe. So, we need to investigate the two signals further to understand how and where they come from. (In the pre-Belfort days, L45HA_T and L45HT_T were signals that came from fire detectors (high-temperature switches) mounted in the Accessory- or Turbine Compartments, respectively. But, in the Belfort era, we have to investigate every signal in detail--and while it would be nice if we could trust the signal descriptions in ToolboxST, as we've seen--we need to "trust but verify.")

<b>l4ct</b> is in lower-case alpha characters, so we can presume that it is from a discrete input to the Mark VIe. A '4' device is a master protective device, as in L4. 'CT' stands for Customer Trip. The signal is a logic "1" when the Customer Trip input is tripping the turbine.

Usually this discrete input comes from the Customer's DCS or some other control system(s)--a circuit that is closed to allow turbine operation, and opens to trip the turbine. This is a VERY important signal to investigate at your site--as it can be from a single "device" or a single contact that has multiple conditions. And, as we've seen with l3cp in the Ready to Start application code, sometimes Belfort just puts a wire jumper on the discrete input terminals.

In the past this discrete input was connected to a normal discrete input terminal board to <Q>. However, I have a sneaking suspicion that it's connected to one of the TREG card Trip Inputs (similar to l86tgt1)--and if not, I don't understand why not (again, Belfort is consistently inconsistent). So, please investigate this input further. If it's a real discrete input (to either a normal discrete input terminal board--or to the TREG card) then you need to do some more investigation to understand where the signal/circuit comes from in the plant, and what causes the signal to change state. Also, if it's connected to a normal discrete input terminal board, I suspect it will be Inverted--but that's just a guess at this point.

So, you have some work to do to finish up L4PSTX2.

I'm learning a lot--as I usually do when I try to teach or explain. And, of course, Belfort keeps it "interesting" with their consistently inconsistent logic/application code and their use of <b>false</b> and <b>true</b> permissives/contacts.

 

CSA:

Your NO/NC explanation helps! I think i should keep that when the NO/NO contact is actuated it will change state in mind.

Yes,i find there is gas fuel portion of L2SFT.
<pre>
L60FSGH L28FDY L2FSTG
------||----------||-----------( )
L2FSTG L86MR1 |
------||----------|/|----- </pre>
L28FDY = Time delayed loss of flame
L60FSGH = startup gas fuel stroke high

So, when the flame is lost and SRV feedback is open past certain point, then L2FSTG will go to logic 1.<pre>
L60FFLH L14HA L2WX L2SFTL
--------||-------|/|--------|/|--------( )
2s</pre>
L60FFLH = Liquid Fuel Flow High
L14HA = HP accelerating speed signal
L2WX = Turbine Warmup Complete, Increase Fuel

I found there is a difference between L2FSTG and L2SFTL:

1) L2FSTL has a timer while L2FSTG doesn't;
2) L2FSTG has a master reset while L60FSGH doesn't;
3) There is L14HA and L2WX in the L2FDTL rung, while L27FDY in the L2SFTL;

As to L45HA_T and L45HT_T,take L45HA_T for example:
There are 45HA1,45HA2 and 45HA3 hazardous gas sensors directly connected to the Mark VI, and the related DI is l45ha1hh, l45ha2hh, l45ha3hh. When two of them is logic 1, then L45HA_T will go to logic 1.

L14H_ZSPD:
"This bit of logic compares the VPRO speed signals to the <Q> (<R>, <S> & <T>) speed signals, and generates a trip if there is a disagreement by more than one processor (<R>, <S> & <T> with respect to <X>, <Y> & <Z>--which are the names of the three independent VPRO cards in <P>)."

CSA, i am not clear that how are speed signals are processed. I think there are six speed signals (1, 2, 3, 4, 5, 6), three for <Q>, three for <R>.

1---<R>

2---<S>

3---<T>

4---<X>

5---<Y>

6---<Z>

So 1 is compared to 4, 2 is compared to 5, 3 is compared to 6, and if two of three disagrees (how is this 2 of 3 logic achieved?), then L14H_ZSPD will go to logic 1.

I am sure not sure whether am i is clear? And i am sure there are misunderstanding.

As to l4ct, again, most of your guess is right! This signal comes from DCS, and it is connected to normal DI terminal board, and it is inverted. But my colleague told me that it should be connected to
one of the TREG card.

As you said the circuit of l4cp is normally closed, and DI connected to TREG can't be inverted. so if it is connected to TREG card, it should not be NC in L4PST2.

Is it right?
-------------------------------------------------------------------------------------------------------------------------------------------------------------------
<pre>
L12HF_ALM L12HFD_C_ALM L12HFD_P_ALM L12H_P 12H_ACC L4PSTX3
--------|/|--------|/|--------|/|---------|/|--------|/|---------( )</pre>

L12HF_ALM = CONTROL SPEED SIGNAL LOSS - HP
L12HFD_C_ALM = CONTROL SPEED SIGNAL TROUBLE
L12HFD_P_ALM = PROTECTIVE SPEED SIGNAL TROUBLE
L12H_P = VPRO HP Overspeed Trip
L12H_ACC = VPRO - Loss Of Protective Speed Signal

I would like to investigate L4PSTX3 next time.

Always best regards!
Neo

 

CSA

I think you may miss this reply.

Best regards!
Neo

 

CSA,

Your response is much awaited. Your replies are very valuable to many of us. it helps in understanding the underlying philosophy and history of GE-design heavy duty Gas Turbine control system.

Please provide us the opportunity to learn from you.

SB

 

Neo,

Are you back?

 

CSA

I am back, and waiting for your reply

best regards!
Neo

 

Neo,

This thread is getting very long--and very disorganized....

<b>L2SFTG</b><pre>
L60FSGH L28FDY L2FSTG

------||----------||-----------( )
L2FSTG L86MR1 |
------||----------|/|-----</pre>
L28FDY = Time delayed loss of flame
L60FSGH = startup gas fuel stroke high

Let's go back to the L4, L4Y & L4Z logic signal/suffix example. L4 is the main, or "namesake" signal for this group. L4Y is a logic "1" some time AFTER L4 is a logic "0", and L4Z is a logic "1" some time after L4 is a logic "1". <b><i>MOST</b></i> logic signals with a suffix of Y are inverse time delay signals--they are a logic "1" some time after the main, or namesake, signal is a logic "0." <b><i>MOST</b></i> logic signals with a suffix of Z are time delay signals--they are a logic "1" some time after the main, or namesake, signal is a logic "1."

I've never really understood why the designers used an inverse time delay L28FDY in L2SFTG because L2SFTG is checking for "excessive gas fuel flow" (and we'll get to that bit in a minute) during firing (starting). L28FD must be a logic "0" during starting (as we've seen from the Start-Check permissive discussion)--so L28FDY will be a logic "1" some time after flame was lost, either after a turbine trip or after a normal shutdown. In other words, flame will have been lost for a very long time before the turbine can be re-started, so I've never quite understood why L28FDY was used; a NC L28FD could have been used with the same results--because this 'permissive' is only saying, "Check for "excessive gas fuel flow" only when there is NO flame, and once flame has been established stop checking for "excessive gas fuel flow.""

Now--about that "excessive gas fuel flow" bit.... As you correctly noted, L60FSGH is only looking at SRV position--<b>NOT</b> gas fuel flow. The assumption is that there <i>will be</i> excessive gas fuel flow if the SRV is above a certain position--but that's an assumption. I've seen many sites forget to open a manual isolation valve upstream of the SRV when starting the turbine and get this alarm--because when there is little or no gas fuel pressure upstream of the SRV then the SRV will open very far trying to make actual P2 pressure equal to the P2 pressure reference. So, the alarm text message is very misleading because it's not actually looking at gas fuel flow--it's only looking at SRV position and assuming that if the SRV position is higher than a Control Constant value that gas fuel flow will be higher than it should be. And that is an assumption.

I've also seen several sites that have low gas fuel pressure upstream of the SRV until there is flame in the gas turbine (mandated by local regulations!) so the SRV occasionally goes "high" trying to make P2 equal to the P2 pressure reference--and this alarm/trip is annunciated. (Remember: When L2SFTG goes to a logic "1" L4PSTX2 picks up, which picks up L4T and trips the turbine by dropping out L4.)

It's been the very rare occasion that the SRV has been unable to control P2 pressure during starting and gone too far open actually causing gas fuel flow to be excessive. (And with most new GE turbine control systems, there is P2 pressure test during purging to check that the SRV can actually control P2 pressure before opening the GCV and sending gas fuel to the combustors.)

The point of this is: Some alarm/trip text messages can be very misleading. One MUST take the time to read and understand the application code (logic; sequencing) for any alarm to be able to properly understand what the alarm message is trying to indicate--and it's also necessary to understand P&IDs (Piping Schematics) to know how the components work and what can happen under abnormal situations. The sites that had forgotten to open the gas fuel isolation valves prior to the start and got this alarm/trip were very angry--and embarrassed--when they found out they'd left the valve closed and had been chasing the wrong condition/sensor (calibrating and re-calibrating and re-re-calibrating the gas fuel flow dp sensors!).

Okay; enough about L2SFTG and it's misleading alarm text message.

<b>L2SFTL</b><pre>
L60FFLH L14HA L2WX L2SFTL
--------||-------|/|--------|/|--------( )
2s</pre>
You wrote:

> I found there is a difference between
> L2FSTG and L2SFTL:

> 1) L2FSTL has a timer while L2FSTG doesn't;
> 2) L2FSTG has a master reset while L60FSGH doesn't;
> 3) There is L14HA and L2WX in theL2SFTL rung, while L28FDY in the L2SFTL

All good catches; good on you! First, the presence of a timer on L2SFTL and the absence of a timer on L2SFTG. A "2" device is, by definition, a timer. There's only one reason I can explain why there's a timer on high start-up liquid fuel flow and not on high start-up gas fuel flow is that someone felt that because of common liquid fuel supply pressure fluctuations during firing which can erroneously cause high liquid fuel flows that a timer was appropriate to prevent nuisance trips during starting on liquid fuel. Many times the liquid fuel supply (forwarding) system is not in GE's scope of supply, and if the system is not properly designed or configured then large pressure/flow fluctuations can occur and GE always gets the blame for these kinds of problems, so it's likely that in order to avoid these kinds of problems (and to avoid forcing the supplier of the liquid fuel supply system to fix their system) someone in GE just decided to put a two-second timer on the signal to avoid nuisance trips and blame and finger-pointing.

To my mind there's no difference between too much liquid fuel during starting and too much gas fuel during starting--they're both equally dangerous conditions. But, since gas fuel flow isn't actually being measured, ..., well, ..., anyway, some things will always be a mystery. I do completely understand the whole liquid fuel supply pressure problem thing, though, having been in the unenviable position of having to tell several project managers that the systems were inadequate--and being told that the problem was GE's problem and that GE had to fix the problem.

The L14HA has been kind of a mystery to me, also, because it usually goes to a logic "1" on a GE-design Frame 6B heavy duty gas turbine at 50% speed, and L2WX usually goes to a logic "1" at about 20-30% speed. These two permissives allow high liquid fuel flow to be detected until the turbine warm-up time is complete--in other words, after flame has been detected; but not after the warm-up is complete, and in no scenario after 50% speed (which is kind of useless).

Enough about these; again, some things will forever remain a mystery and have little or no effect on normal turbine operation.

As for speed signals, I dislike it when you say:

> I think there are six speed signals (1, 2, 3, 4, 5, 6),
> three for <Q>, three for <R>.

You should be CERTAIN, and to be certain, you need to look at the documentation and drawings provided with the turbine and control system. For most new turbines (new since about 1990) with TMR control systems there are, indeed, six shaft speed pick-ups (for a single-shaft turbine). 77NH-1, -2 & -3 are usually connected to <R>, <S> & <T>, respectively, and 77NT-1, -2 & -3 are usually connected to <X>, <Y> & <Z>, respectively. They are all identical speed pick-ups, three used for "control" (and Primary Overspeed Detection), and three used for overspeed detection (Emergency, or Back-up, Overspeed Detection).

Speed detection is kind of a special case in TMR panels, and how it's done is usually dependent on the version of the panel. But, you're basically correct--each control processor checks it's speed signal against it's associated Protective Processor's speed signal, and if two control processors and their Protective Processors disagree, then there's a trip because of the disagreement.

The differences in how speed signals are processed between Mark IV, Mark V, Mark VI and Mark VIe would make your head SPIN VERY FAST! (It does mine whenever I have to try to explain it because it's all kind of a background thing which one should never have to worry about--and it got more reliable with each newer version of Speedtronic.)

As for the l4ct input, I would argue that it should have been connected to the TREG card and not the l86tgt input. But, you are right; it's contact sense would have to be changed--and that would also make "reading" the signal from its name problematic, also (like the l86tgt signal confusion!!!). But, I never recommend changing signal assignments without GE's approval--especially critical signals like this.

Anyway, that's all the time I have for today. More later! Good to have you back!

 

CSA:

I opened a new thread for this one is a little long and disorganized.

http://www.control.com/thread/1409237312

Best regards!
Neo