SCADA for gas project .. asking for ur help.

[DeltaX]

Dear all,

I'm a little bit new in this feild, so i need ur advice for starting run a project for SCADA gas project.

first of all, Is there a specific points should be taken in consideration for gas projects that not apply for power line projects?

from my basic research on the web and some books;

there are two basic modes of capture of input data which maybe used by the central processing facility of the SCADA:

• Scheduled capture, whereby the local units are polled on a regular basis and all input data are transferred; or
• change of state capture,whereby only input data which have changed are transferred.

could you plz calrify this for my and which one you recommended for design

anothe question; for the I/O network there are two options:

• Either all RTU are connected on the PLC remote I/O communication network and the SCADA servers connected via another link to the PLC that as that as the role of interface;
• or all the RTU and PLC are directly connected to the SCADA servers via the I/O communication network

which one from your experience you recommended.

Your help is highly appreciated.

 

[PLCMentor]

First, I cant respond to gas versus power line - so I wont.

The two modes of capture that you mention I have seen to be more associated with data logging. In one method (and the oldest) data is logged at a set frequency (ie, one minute or 10 seconds). State of change capture, as you term it, is more efficient in that it only logs data points that change. It works like this: you set up a deadband or a specific amount of change that is meaningful to your application. For a temperature it may be 1 degree for a flow it may be .01 GPM. The system will only log points for data that changes equal or greater than the amount of your dead band. So for our temperature example if the temperature is 21C then at 21.5C no point will be logged. At 22C a new point will be logged. This is efficient because it only logs data that has changed by an amount that is determined important by you. For a pressure signal you could set your deadband so that it would log lots of data points on a spike. This would allow you to better diagnose the problem. Generally there is a maximum time period between logged points that you can set so that it will log a point every 30 minutes reguardless of the deadband.

I understand that I am descibing this in the terms of logging, but that is where I have seen this used. I would assume it would work similarly if you needed to use if for collection.

As for your I/O network I would prefer all I/O to come into the PLC. This gives maximum flexibility over how this information can be used. I assume you would be interested in only monitoring information from RTU's. I would never control the I/O directly from the SCADA. I am a firm believer in PC's not being used for control. They have been designed for the desktop and should stay there. PLC's have been designed for real time control and are well suited. PC's are great for gathering information and for allowing that 'window' into your PLC.

Russell

 

[BobB]

Most SCADA systems can be programmed for scheduled scans but generally scan all the time.
Change of state (COS) is a different matter. I have used OPC servers that scan change of state but I think the devices have to have the ability to indicate (transmit) change of state.
For example, some Device Net I/O has change of state built in but most I have used are just scanned.
Change of state and scheduled scan is good for datalogging, as mentioned, but generally I find regular scan is the best method. When one considers that I have a Citect system running through Ethernet and then drilling into another peer to peer network of 5 PLCs and Citect reports over 15,000 digital reads and 180 word reads per second, unless the job is huge I can see no point in COS.
Some BMS (building management systems) use COS, particularly the BACnet ones, but generally their networks are pathetically slow, the RTUs or PLCs atre pathetically slow and quite often fall over and that is the only way they can get a reasonable response.
Further, BMS systems are used with air habdling etc and the response times are relatively slow and the slow networks and RTU/PLCs are OK for the job.

 

[DeltaX]

I'd like to thank you first for your response Russell.

This is a SCADA system project which should be installed for the branch line gas to power plant.

I would never control the I/O directly from the SCADA

From my understanding "Plz correct me if I was rong", the SCADA system should perform all supervision, control and data acquision tasks.

The SCADA system should collect and transmit all the necessary information from and to the feild through Remote Terminal Unit"RTU" and/or programmable logical controlar PLC via a fiber optical network all along the pipeline.all the data should be gatherd in the SCADA servers installed in the control room and be monitored monitored from operator station lcated in the same room.

how can that match with what you have mentioned above?

anothe question if you don't mind;

for specific requirments with regard to amount of data, bandwidth, response time and longest distannce for large area a DNP 3.0 and IEC 870-5-101 can be used.

DNP 3.0 sends small number of large sized data while IEC 870-5-101 send a large number of small sized data. my confusion comes from this artcle:

"The choice of the protocol between options is made depending upon what country the target application will be in, i.e in North America DNP 3 would be the obvious choice, while in Europe IEC 60870 would be the clear winner"

what your advice for using the DNP or IEC. what is the factors that affect to choose one of these?
Is that technical issu or commercial issue?

as you know the cost is an issue, especially in the present economic situation.

Can the SCADA system web-enabled or not? If yes how guarantee the real-time??

Thanks in advance​

​

Omar Bukra

 

[Maj. Toxido]

anothe question if you don't mind;

for specific requirments with regard to amount of data, bandwidth, response time and longest distannce for large area a DNP 3.0 and IEC 870-5-101 can be used.

DNP 3.0 sends small number of large sized data while IEC 870-5-101 send a large number of small sized data. my confusion comes from this artcle:

"The choice of the protocol between options is made depending upon what country the target application will be in, i.e in North America DNP 3 would be the obvious choice, while in Europe IEC 60870 would be the clear winner"

what your advice for using the DNP or IEC. what is the factors that affect to choose one of these?
Is that technical issu or commercial issue?

as you know the cost is an issue, especially in the present economic situation.

Can the SCADA system web-enabled or not? If yes how guarantee the real-time??

Thanks in advance​

Omar Bukra

Please see the link http://www.plctalk.net/qanda/showthread.php?t=41397&highlight=iec+60870-5-101

 

[DeltaX]

Thank you Maj for ur help.

what about scada system web enabled?

 

[Maj. Toxido]

SCADA Web Enable

Thank you Maj for ur help.

what about scada system web enabled?

Yes there is. Please find IEEE C37.1 Standards for SCADA and Automation System.

 

[PLCMentor]

how can that match with what you have mentioned above?

You will have to forgive me as I completely misunderstood your use of RTU. Your were writing about Remote terminals and I was thinking remote I/O. So, to be clear, I would never control the I/O directly from the SCADA system. Actually to be clear I would want the PLC to control the system and relegate the SCADA to being a window into the system and a data collection medium. I do not put control algorithims in my SCADA development. I do allow operator control from the SCADA.

Russell