PLC

[durussel78]

Heya Guys
Question
How come you dont see more PLC's in buildings?
I understand price is about triple to have a system like this in there but from what little I know and even less than Ive read these things are like cockroaches - they never die?
One guy put it best - He asked have you ever seen a HVAC Controller 2 miles beneath the earth in a extremelly humid and dusty environment.
The answer is no - he goes on to explain all of the things in a coal mine PLC's control. Other than a Siemens system at the mall I have never seen a PLC used in HVACR. Everything was HVACR specific?
Has anyone ever been the architect behind the design for one of these systems?

[BACnet]

Equipment for coal mines has to be 100% spark proof. The types of relays we use above ground could cause an explosion down in a mine. That is why every piece of electronics in a mine is specially designed and specially certified.

[durussel78]

Heya
This dude was just comparing robustness.
Actually one of the points of robustness.

[BACnet]

People post here all the time about HVAC systems put in place 25+ or more years ago that are still chugging along just fine. There is big business in gateways to speak to those older systems.

Reading all of those posts I don't get the feeling that there is any categorical lack of "robustness" in the HVAC industry.

But I would prefer to use an HVAC controller to a PLC when doing HVAC work because they were designed to make HVAC work easy. They have communication lines to allow them to communicate over the net, they have applications to control common HVAC equipment, they have control loops that are easy to tune for the time constants commonly found in HVAC applications, etc.

That PLC-based mall system you alluded to will be a real bastard when they renovate or build on to the buildings. On the other hand, if they had chosen a proper system (Lon or a BACnet) adding to or modifying it will be a piece of cake.

[only1doug]

PLC: Programmable Logic Control isn't often used in buildings as it isn't very well suited to analogue applications such as temperature, humidity and pressure.

[durussel78]

So I see.
The mall system is somewhat 7 years old.

I have used quite a few hvac specific controllers - easy to understand, robust, easy to tune, and easy to adapt to....
Watching how its made and saw all those machines working in sequence with one another and not missing a beat just sparked my curiosity into PLC.

Trane try to adopt to that mall system and had a pain in the a$$ time doing it.

The other thing that drew my interest is the fact of picking up a ACD book and ordering whatever you needed versus dealing with ALC or Trane or Novar and having to locate a dealer, pay there percentage for the equipment, and pay for programming if they held the licensing for the system.
Just trying to find a way around the man!

[sysint]

Well, we have a controller that isn't licensed that programs with IEC61131-3. So, if you like ladder logic you can use it. People typically like to use functional block when available.

Program cycle times down to 10ms.
No reboot on program changes.
Run multiple program instances.
Offline and full online debugging.
100% backup capability to a SQL DB.
Redundant power connections.
Up to 320 physical I/O with each module having backlit LCD display.

...anyway, that's some features more "PLC-like". And they can speak BACnet/LON/Modbus/OPC and have alarming, scheduling and trending built in. Scheduling is another feature that usually PLC guys don't care about besides open protocols. So, usually with PLC in building automation they fall short with many features you take for granted in building automation systems.

Really, all the devices are PLC's. Some just have better features and construction.

[dapper]

We recently installed a small project where the customer insisted on PLCs. I had a new kid that was familiar with the programming so we agreed to do it. The costs were staggering to get add-on modules for temperature inputs and analogue outputs. In fact for 2ins or 2outs the cost was almost equivalent to the base module and we could not get combination temp and analogue ins so if we had one temp sensor and one 0-10v in then we needed two modules. Any more than two then we needed an additional module.

Cost wasn't bad if you are only doing digital ins and outs but it seems like the analogue thing is a whole new world to them and costs get real crazy. I will never consider using a plc again in an HVAC environment unless the customer is first made fully aware of the costs differences involved.

[drjfjr]

I have several pieces of equipment that used PLC's as the temperature controller as well as control of the automation. The PID loop programing in a PLC is a real pain. To tune and maintain temperatures with constant changes in the external variables was a constant challange. Had a lot of over shoot , static discharge spikes and ground loop problems that cause temperature faults. Put in some off the self temperature controllers set them and problems disappeared. The PLC runs the machine and the temperature of the pads is nice and steady. The Algorithms of dedicated temperature control equipment tend to be better since each generation has enhancement where as a PLC's advantage is that it is simple to change i/o and sequences and is very adaptable. Some of the newer plcs have higher language such as C++ and others that allow them to work calculations better than the old ladder logic only PLCs so one day the temperature programs may be canned and allow the PLC's to control temperature easier.

[orion242]

How many times have you seen a Cray super computer running a simple calculator?

Better yet, how many PLCs have full blown scheduling built in?

[davem]

The only time I've had to program a PLC for anything related to temperature control was a recent datacenter we did, where the chiller plant was considered crucial enough to require redundant CPU's - none of our DDC solutions offered true redundancy, so we went with a PLC brand that offered both redundant power supplies and redundant CPU's.

I'd never seen such a cool setup - if one of the two CPU base modules takes a dump for some reason, the backup picks up the program where the primary left off - within a millisecond or two, with no hiccup in the outputs. Latching relays offered some additional protection from hiccups, but flicking back and forth between the CPUs without a glitch was neat to see.

[crab master]

Which PLC brand?

[sysint]

The only time I've had to program a PLC for anything related to temperature control was a recent datacenter we did,... - within a millisecond or two, with no hiccup in the outputs. Latching relays offered some additional protection from hiccups, but flicking back and forth between the CPUs without a glitch was neat to see.We have ability for redundant power but when I see this you could do this with two separated controllers... probably still less money. No doubt, very slick.

[davem]

It's a brand called KDT Systems, from Korea (their CIMON-PLC line)...a bit tough for tech support when they get into the office once I'm already home, but they definitely have a nice product (once you get past the Korenglish documentation...).

I hadn't played with PLC's since electrical school (ABB SLC-500's...), but I found this line fairly painless to work with (lots and lots of documentation...) And, since it talked MODbus right out of the box (I don't think that this is unique among PLC's, but certainly handy to have), it integrated cleanly into our Tridium front-end running the rest of the datacenter.

[printer2]

Have not played with PLC's for a few years but can say they are capable of doing everything a dedicated HVAC controller can. Temperature, pressure, humidity, no problem just have to get an analog input card that accepts the sensors you want to use. But why go through the trouble and expense when the HVAC industry has products specifically designed for the application?

My brother designs heating/ventilation equipment for the farming sector. Why not just use regular HVAC controls? The market is even lower cost than the HVAC market and because of it their equipment is designed with that in mind. Each tool has its own features and limitations that have evolved for each niche.

[jam2k5]

Hi !
In our company we use PLC to control lights in buildings,we make some remote I/O island , is cheaper , also we use PLC in critical facilities such as farmaceutical and industrial HVAC.
Normally we use S7 or Premium critical facilities (redundat) and Beckhoff for the non-critical .

Price of low-end PLC is cheaper than some HVAC dedicated products and have tons of comunication options like Modbus,IP, OPC, LON, even Bacnet...

[sysint]

Hello jam2k5, I need to disagree because there are building automation controllers with good power, multiple protocols and importantly better capability with AST (alarm, schedule trend) and built in visualization that caters to the buildings market. Further, I find many bacnet drivers for PLC lacking in features with priorities, objects, objects supported, and restrictions on client mapping or server objects. Oftentimes there is no bacnet scheduler, but some internal scheduler.

For instance we have a controller that supports many PLC type features including on board OPC server, BACnet, LON, Modbus, Mbus with full AST, visualization, backup, down to 10ms cycle times, IEC program interface to 320 I/O. So, very powerful by most standards in building automation.

So, it may be the device has a driver for another language, but that is just the beginning.

[jam2k5]

Hi sysint , i agree with you , there are controllers with lot of power and capability , but the main idea is that we use low end Plc cause money talks , they are cheap and have a flexibility in the configuration of I/O that most controllers don't , i can have the mixture of signals that fits perfectly with my requirements and still have room for spares.

Normally we use Plc only in light signals or alarms collected from 3ºparty devices, as they are binary data and theres not much strategy behind them, AST is done by Scada or by some centralized Plc with redundancy.
We do not use analog with Plc cause are pricy , for pure hvac we use controllers such as tac or pxc ...

That's the way we do it, not the way i want to do it :-(

[sysint]

You must be talking DIRT cheap.

[wolfdog]

They have been using PLCs in printing press chillers for years.
http://minshewmechanical.com/images/grafix5vo8.jpg

[sysint]

Of course. But now you can buy a LINX and get like performance and also nice features like LCD backlit display on the I/O.... and it's in a European form factor design so it fits nice in their electrical control cabinets.... looks like an extremely clean controls cabinet. Is everything else in the room spotless?

[jam2k5]

Think in a setup with 36 DI and 32 D0, if you make the setup with Linx121 plus LIOb 101/102 you get 5 times the price of doing it with Beckhoff BC9100 plus its I/O.

So if you need performance, maybe a mix of two world would be the best , a Linx 121 with a BC9100 and I/O communicating via Modbus/TCP ?

[Knife Switch]

I worked for a while at an electrical company that was branching into HVAC, they were traditionally automation guys, but alot of the automation work was drying up.

Anyway, as they were the preferred guys for alan bradley plc's they had the cheapest options locally, and AB's were looking to get into the hvac market too. Plus, the name AB makes consultants happy, it's known as a good brand and the most common one I've seen when a HVAC job has to use PLC's. I thought "awesome", we were gonna do well.

Some of the problems were no firmware, and the physical differences between HVAC and automation. Automation guys like their controllers to be centralized, in one (or more) boards (which are built to a HIGH standard, I was really impressed) rather than the more traditional HVAC way of "scattered at every VAV box/FCU in the roofspace". Thus, their controllers hadn't really needed to evolve to include the TEC/SPC/Small end of line type controller that HVAC has. Don't get me wrong, I WISH the HVAC world was alot more like the automation world (and some jobs are) but the "small, dirt cheap , HVAC firmware loaded" controller just doesn't exist for them. They used to make fun of the cheap HVAC controllers, comparing them to theirs, but I used to counter with "That one's $200, the other is $5000"

In the end, it just wasn't financially possible to compete with problems like that.

BTW the range of "add-on's" that they have access to would blow your mind. Absolutely awesome stuff, like proximity detectors that distinguish between metal and people, light curtains that do all kinds of crap, and a s*&^load of stuff that I didn't even see. (And that's 99% useless to us).

Great support too, the manufacturers reps had a tonne of stuff to lend you for fault finding (like network analysis stuff worth 10k).

But for six figures worth of hardware, you'd hope so.

[sysint]

Think in a setup with 36 DI and 32 D0, if you make the setup with Linx121 plus LIOb 101/102 you get 5 times the price of doing it with Beckhoff BC9100 plus its I/O.

So if you need performance, maybe a mix of two world would be the best , a Linx 121 with a BC9100 and I/O communicating via Modbus/TCP ?Well, I cannot verify your price concerns here plus you are in a different area.

However, I suppose with your specific instance that could be true in some ways because the current I/O is tailored to equipment. Unlikely you can get 36DI and 32DO without getting some AO and AI. And, we don't like pushpin connectors, you get the display(s), the hot swap... What I'm saying is maybe there are more issues of viable difference than you see "face value".

Yes, you can use a 121 and use Modbus RTU/TCP or "cheapie" LON I/O. And, yes you can use a very inexpensive 110/111 and do the same. Or, a 110/111 and that nice LIOB I/O with backlit display.

Really a better comparison is a 110 with LIOB. Actually, we still come out ahead because the LIOB module is better with the backlit display.

But finally features come at a price. And, compared to labor and soft costs isn't as much a consideration on projects of any size.

[jam2k5]

But finally features come at a price. And, compared to labor and soft costs isn't as much a consideration on projects of any size.

That's the key ... try to make understand sales people that , is our battle every day. The cost of Plc programming are very high, and are only affordable if you have a very standarized way of doing things , everything else is quite a nightmare ...

Up to date i havn't make two projects with the same hard / configuration, i wish some loytec (not only L-IP) fall in my hands ...

[sysint]

I would like that too so contact our sales (I am in America) and let me direct you a proper direction. Thanks for your input.

http://www.ifema.es/ferias/climatizacion/default_i.html

March 1 – 4, 2011
Feria de Madrid - Hall 10, Booth 10A09
Madrid, Spain

[jam2k5]

Thanks for the appointment ...

i've already contacted loytec sales here in Barcelona (e-controls) that's why i know your products and prices.

But know in Spain , things are going quite badly , finnacially its quite imposible to make an offer with this type of controlllers , people are just trying to keep systems on with a small budget , so the only thing that fits ( initially ) are mixed solutions, Plc / controllers.

[EMPadvocate]

I use the Automation Direct DL06 for central plant work using Turbocors and VSDs. I use the com2 port for direct compressor control and the ECOM100 card for modbusTCP talking to a JACE.. For the VSD comm I use a modbusTCP gateway (moxa). Once you get the PLC code right it just works.. The PLC with the ECOM100 card is @ 430 USD.. other cards are inexpensive. Documentation is excellent.. I wish they had a IEC1131-3 programming solution though like the Loytec product. I saw the Loytec at AHR and I feel it is the most advanced HVAC solution out with everything a controller should have (VNC, SQL, IEC1131-3, web server, I/O LCD) but very spendy...

[sysint]

There is some inherent long term cost advantages to not need multiple controller styles in your systems along with management controls in place. This adds cost to the front side.

Possibly also you are indicating the need for a smaller controller with some of the same features with directly connected I/O that more closely related to the base DL06?

[EMPadvocate]

The base DL06 has 3 slots open for analog I/O since one is used for the ECOM100 ethernet device. You can add another modbusTCP device in the ECOM100 configuration or another DL06 or a gateway and then it is peer to peer; there are a lot of modbusTCP I/0 devices on the market - you also have the option of putting a modbusTCP card in a drive but I just use a gateway.

[sysint]

I would like to discuss this more directly with you. You bring up some viable points.