PDA

View Full Version : Variable Frequency Drive Reciprocating Compressor.



kothari
01-02-2008, 05:22 PM
Seeking help from forum regarding installation of VFD(Variable Frequency Drive) on a semi seal refrigeration Copland Reciprocating (hp 15) compressor.Reason are to conserve power and to reduce vibration while loading and unloading the unit.

Kothari:cool:

Samarjit Sen
03-02-2008, 05:15 PM
What little I have learnt about VFD is, that the compressor should have an oil pump. You can operate the compressor with low rpm by changing the Frequency. I do not know about the vibrations, but yes the compressor would start at a lower rpm and would be able to manipulate the capacity as per the requirement by reducing or increasing the rpm/ frequency.

Josip
03-02-2008, 05:22 PM
Hi, kothari :)


Seeking help from forum regarding installation of VFD(Variable Frequency Drive) on a semi seal refrigeration Copland Reciprocating (hp 15) compressor.Reason are to conserve power and to reduce vibration while loading and unloading the unit.

Kothari:cool:

agree with you, but first try to contact Copeland site

http://www.ecopeland.com/

and see what is recommended or not....

Best regards, Josip :)

SteinarN
23-02-2008, 12:26 PM
I have installed several systems with Danfoss VFD and Copeland Discus compressors. It works wonderfull. :)

The compressors i have used is 3 cyl for fresh and frozen R404A 50Hz motor. Acording to Copeland, the Hz limits are 25Hz to 60Hz and for short periods up to 65Hz. I run them between 23 and 67 Hz. Important issues to consider is the size of the VFD. It should have a continous output amps capacity of at least 60 to 70% in excess of the max amps on the nameplate of the compressor if it is an 3 cyl comp. If its a 2 cyl comp, select a VFD rated at least twice the motor amps. For 4 cyl and larger comps you could probably get away with a VFD rated less than 60-70% above the comp motor. This is bc of the high torque needed to overcome the torque pulses when starting a piston compresser. The torque pulses get bigger when the comp has fewer cylinders.

One more issue is that the VFD MUST have frequency bypass function. There is a considerably possibility that you will get some serious vibrations at certain Hz/rpm. You must be able to confugure the VFD to skip those Hz.

The Danfoss VFD i installed has worked flawlessly after i got it set up properly. There wasnt any experiences at the seller of those VFD so i had to figure it out by my self.

One problem with comps is that they dont state the true power of the motor at the rated amps and voltage on the nameplate. In setting up the VFD you must configure max amps, nominal Hz, voltage and the power of the motor.

Do NOT enter the Hp of the motor converted to kW in the power data to the VFD. This HP is an almost randomly selected value. Try to take the voltage, (minimum voltage if it is lets say 420-460V) multiply it with the max amps, multiply again with 1,73 (since it's a 3-phase motor) and multiply that again with 0,85, (to correct for the reactive power) and finally multiply with 0,82-0,90. (0,82 for small comps, and 0,9 for large comps, correctionfactor for the efficiency of the motor) The sum should be in the proximity of the true power in W of the electric motor. Divide this value by 1000 and you have the kW rating of the motor, enter it in the VFD. You also have to run the auto optimization function on the VFD to let it measure some important electrical properties of the comp motor.

When this is done, the VFD is set up to fit the motor as good as possible.

One more issue is to configure the VFD to deliver highest possible starting torque to be able to start the comp and to avoid installing a disproportionately big VFD. I got in serious trouble before i figured out how to do it. At my first system, the VFD wasn't always able to start the comp and it tripped several times the first mont with total lack of cooling as a result. Got a happy end user then. :o

There is several configurations to set for the starting torque. On Danfoss VFD there is a configuration, number 110, "motormagnetization at 0HZ", which is set at 100% as standard. This is a very important configuration to set in order to acheive highest possible starting torque. When i increased this setting to 200% all my startingproblems and tripps with nasty alarms disapeared. :cool:

This is some of the most important issues to be aware of and consider when adapting a VFD to a compressor. All the other issues and configurations nessecary for the regulation of the system is not covered in this brief explanaition.

Electrocoolman
23-02-2008, 02:17 PM
I've also used VFDs, not for refrigeration compressors but for hydraulic pumps.
There are different types of VFD on the market these days, which are designed for a range of different applications.
These VFDs have different software inside, different V-F curves and are suitable for Constant-Torque loads and give a much higher starting torque capability. You do not then need to oversize the drive or do an unloaded start.

The motor can be ramped up to speed over a number of seconds - This removes any start up and any loading up surges.
The rundown can be likewise controlled.

As Steinar says in his excellent post, you need to enter the motor parameters carefully. The units I have used (Telemechanique - ALTIVAR units) make it simple; you then carry out an Autotune to allow the drive to measure the winding characteristics.
It uses these for motor protection, but I don't know if refrigerant cooled windings would upset the drives calculations for winding temperatures etc.
It would possibly be better if your motor has built in winding protection thermistors (as a lot of compressors do)
On the other hand, refrigerant cooled windings would allow enhanced low speed operation compared to a standard aircooled motor, which runs hotter at low speed due to inefficient air cooling.

VSDs give you a much higher motor protection compared to standard contactor/overload combinations, but heed advice and take note of max and min speeds, and lubrication considerations.

Don't forget that VSDs need ventilation...some plant room locations can be on the hot side.

You also need to consider your control side carefully and what parameter(s) you are going to use to vary the speed.

SteinarN
23-02-2008, 04:25 PM
Hi Electrocoolman :)

I've read other posts from you. You have very good knowlegde about electrical issues. Thank you for valuable contribution.

The Danfoss VFD in the VLT5000 series, which i have installed, is sort of a high end VFD. A recips (2-3 cyl) is a very hard task to start for a VFD. Maybe other people in here have experiences with other brands to drive recips. Until then i will strongly advice to specify a VFD with substantial powerreserve. I have no experience with VFD and large recips. (6 and 8 cyl) It is possible, like Electrocoolman says, that it is not necessary with a substantial powerreserve for those instances. However, like Electrocoolman says, be aware of the temperature inside the plant room. At temperatures above some 30-35 C it is necessary to reduce the load on the VFD, ie select a larger VFD. It is also favorable with regard to the lifetime of the VFD to select one with powerreserves.

Electrocoolman mentioned the termic load on the compressormotor. Since the motor is cooled by the always flowing refrigerant, it is not necessary to limit the torque at low rpm. If the VFD has a function for "separat or external cooling fan", then activate this function. Or, if this function doesnt exist, then disable the function "termic motorprotection" in the VFD. If this function is not activated/disabled the VFD cant suply full torque to the compressor at prolonged running at low rpm, but instead the VFD would assume the motor is overheating and thus shut down with an alarm for overheated motor. The max amp setting in the VFD and/or the thermistor in the compressormotor take care of the motorsafety.

Core4 Guy
24-02-2008, 07:45 PM
Guys, this is great information that I am going to check and apply it at one of my new sites Monday. I have 2 Carlyle Compressors that run on VFD's and the only issue I see is that I have to set the VFD low hertz speed above the compressors lowest speed where the compressor can still make oil pressure. I set the starting time as quick at possible to ensure the machine generates enough oil pressure quickly. The is probably not an issue with copeland compressors as they have a different oil pump design than carlyles.

Energy consumption and VFD's. One can easily equate VFD's with energy savings, but this is not always true. VFD's have parasitic losses associated with them and they create inefficiencies in motors. There must be a good reason to install them on a recip before a claim of efficiency is handed out.

I use VFDs to ensure the design mass flow rate of a compressor is not exceeded when I am floating head pressures. I also set the VFDs to operate as close to the slip frequency of the motor (without slipping) as possible to ensure the highest efficiency of the motor. This can save up to 8% at low htz.

VFD's like anything else are not a plug and play energy reducer. The design of the system is the most important thing.

US Iceman
25-02-2008, 02:27 AM
Before everyone gets excited about energy savings using VFD's you might want to consider a recip compressor is different than a fan or pump. The big energy savings on centrifugal fans or pumps is valid with: RPM2/RPM1^3.

This does not apply to recip or screw compressors as they are positive displacement machines. The energy savings when VFD's are used on compressors is a result of better part load capacity control.

So... before you try to say these save a bunch of energy, you need to investigate what you try and how hard you sell it.;)

SteinarN
25-02-2008, 09:37 AM
Where i have applied VFD is on one compressor, centraliced DX systems. On such systems the VFD gives the posibility of significant energy savings. On a system for frozen cases i found i could run at an evaporator temperature of -31,5 oC instead of the suplier specified -35 oC. That was wery simple to adjust seamlessly with a VFD. As Iceman says, it all has to do with capacity regulation. It is there the possible energy savings is.

On large systems with a compressor rack, with maybe different sizes of comps and/or comps with capacity steps on themselves, you normaly have the posibility to a near seamless capacity regulation without VFD. And therefor there is not anything to save.

The same applies to one comp-one evaporator systems, at least if the evaporator have a large capacity compared to the compressor.

Peter_1
25-02-2008, 09:40 AM
Some thoughts: first, USIceman is right, you will see the biggest benefits in a multiple evaporator setup like a pack/rack in a supermarket where the compressors can run in partial load.

The only advantage I see in a single setup is when lower ambient occurs you simultaneously reduce the speed of the compressor to maintain a LP as high as possible.

One of the benefits here in Belgium - and perhaps elsewhere - to use VFD's is the financial help of the Belgium government. They pay you a certain amount for a VFD for each kVA you've installed. (577 € for a 5 kVA, 1047 €for a 10 kVA, 1816 € for a 25 kVA, 2.678 € for a 50 kVA to give you an idea)

Core4Guy, I have no experience with Carlyle but we ran Bitzers for a longer time as low as 10 to 15 HZ as a test and monitored the oil pressure, all equipped with an oil pump. The oil pressure remained the same at 15 Hz like it was at 60 Hz.

We used several times Invertek inverters from the UK, cost far less then a Danfoss and performs just the same.
As with the Danfoss, you can connect direct your LP probe (4-20mA) to the inverter and let control the probe direct the inverter.

Then you have the not so cheap German Kimo inverters, specially made and programmed for refrigeration compressors. You can enter direct the used gas and the desired evaporating pressure in Pa or bars for example.

The "bypass speed" is also called the 'jump-speed'.

SteinarN
25-02-2008, 09:50 AM
Some thoughts: first, USIceman is right, you will see the biggest benefits in a multiple evaporator setup like a pack/rack in a supermarket where the compressors can run in partial load.

The only advantage I see in a single setup is when lower ambient occurs you simultaneously reduce the speed of the compressor to maintain a LP as high as possible.

One of the benefits here in Belgium - and perhaps elsewhere - to use VFD's is the financial help of the Belgium state. They pay you a certain amount for a VFD for each kVA you've installed. (577 € for a 5 kVA, 1047 €for a 10 kVA, 1816 € for a 25 kVA, 2.678 € for a 50 kVA to give you an idea)

Core4Guy, I have no experience with Carlyle but we ran Bitzers for a longer time as low as 10 to 15 HZ as a test and monitored the oil pressure, all equipped with an oil pump. The oil pressure remained the same at 15 Hz like it was at 60 Hz.

We used several times Invertek inverters from the UK, cost far less then a Danfoss and performs just the same.
As with the Danfoss, you can connect direct your LP probe (4-20mA) to the inverter and let control the probe direct the inverter.

Then you have the not so cheap German Kimo inverters, specially made and programmed for refrigeration compressors. You can enter direct the used gas and the desired evaporating pressure in Pa or bars for example.

The "bypass speed" is also called the 'jump-speed'.


I also conected the 4-20mA LP sensor directly to the VFD. Aditionally i was able to configure a lower setpoint where the comp should stop in that case the suction pressure become wery low as the load decreases. I also was able to configure a minimum stop time before the compressor is alowed to start again in order to avoid a almost imediate start after compressor stopp. All this directly in the VFD without an aditional PLC.