VFD versus fluid coupling
Today we open a topic that might polarize our readers:
Comparison between variable frequency drive (VFD) and a fluid (hydrodynamic) coupling.
Of course, this blog is dedicated to VFDs. Nevertheless, we intend to provide an objective benchmark as far as possible. Please note that there is lot of information on the Internet that is not based on facts or physics. Instead, it is based on emotions and manipulated information. At this place we like to remind everyone dealing with such benchmarks that “typical catalog values” are not always the same as project specific values, especially if those values shall be guaranteed.
Fig. 1: Electric versus hydraulic variable speed technology
Variable frequency drive is based on power electronics. The VFD is typically realized as indirect frequency converter (there are also direct frequency converters like cycloconverters, but their use is very rare). Indirect means that the frequency conversion happens in two stages. First, the rectifier converts alternating current into direct current (LCI) or alternating voltage into dc voltage (VSI). The dc bus contains either a smoothing reactor (LCI) giving the VFD a character of current source or it contains set of dc smoothing capacitors (VSI) giving the VFD a character of voltage source. The dc bus also serves as intermediate energy storage and allows exchange of reactive power. The direct current or dc voltage is then converted into alternating current or voltage by means of inverter. Due to the two-stage conversion the output frequency is independent on the input frequency of the grid. VFDs are air-cooled or liquid-cooled depending on power rating, installation or customer requirements. The power density is lower compared to hydraulic VSD. However, the VFD is located in a different room than the motor so space is not necessarily as issue. The ramps for acceleration and deceleration can be set arbitrarily. The motor is inherently soft started and high starting current is completely avoided. The dynamics is adjusted by control parameters and VFDs can be found in variety of industrial areas from low dynamic applications (most of pumps, fans and compressors) up to very high dynamic applications (steel rolling mills, specific test stands). The input power factor is high; either above 0.95 for diode front end drives or even 1.0 for active front end drives.
Fig. 2: Liquid cooled large capacity variable frequency drive
Fluid coupling is based on hydrodynamic principle. It is known for decades and commonly used in cars in a very similar setup (see excellent description in [1] from 1950s). The torque is transmitted through a fluid medium – usually oil. The pump wheel (impeller) is driven directly by the electric motor and the turbine wheel (runner) is driven by fluid circulation. There is no direct connection between the pump wheel and turbine wheel. The turbine turns at lower rpm than the pump wheel and this is the fundamental principle of fluid coupling. The difference between pump speed and turbine speed is the slip of the fluid coupling. The slip is proportional to the fluid level. This is the core of the control – a scoop tube controls the level of the fluid and correspondingly transmitted power and slip between impeller and runner. Fluid coupling is a hydraulic variable speed drive. As such, it reaches high power density and is very compact. It has rugged design that is suitable for both indoor and outdoor installations. From its nature it is quite simple solution and uses just few components. However, don’t forget that besides the fluid coupling there are additional components such as external re-cooler, oil filter, scoop tube for speed control and other. The efficiency is very high when operating just slightly below the nominal (full) speed. Some manufacturers indicate up to 97%. However, when wider speed range is required the efficiency of a fluid coupling suffers and cannot compare with VFD. The fluid coupling inherently provides soft torque transmission and reduces torque spikes on the load side. In a fault scenario it acts as kind of mechanical fuse and protects against overloads and transients. On the other hand, the start of the electric motor is still direct on-line with corresponding large starting current, thermal stress of the motor and voltage sag in the grid. If this is not acceptable, an additional starting system consisting of pony motor and synchronization equipment is required (increased footprint, cost and complexity).
Fig. 3: Hydrodynamic (fluid) coupling
Strengths and weaknesses
Strengths of VFD
– Superior efficiency [2]
– Excellent power scalability
– High dynamics (if needed)
– Full starting torque from zero speed (if needed)
– Inherent soft starting of motor
– Inherent full motor protection (overload, unbalance, over-voltage, phase loss, stalling etc)
– No extra space in machine room
– VFD location independent on motor location
– Multidrive option (footprint, cost and energy savings compared to individual VFDs for each motor)
– High input power factor (> 0.95 for diode front end; 1.0 for active front end)
– Output frequency can be higher than input frequency
– High accuracy of speed control without the need for speed sensor (encoder-less)
– One single VFD can soft start load 1 and afterwards start-up and drive load 2 (e.g. ball mill and SAG mill)
– One single VFD can sequentially start-up multiple motors using synchronous bypass
– Flexible cooling solutions [3]
-direct air-cooled
– air/air heat exchanger
– air/water heat exchanger
– water/water heat exchanger
– Short mean time to repair –> high availability
– Multiple options to further increase availability (redundant cooling, bypass switch etc)
– Straight forward for retrofit projects
– Oil free technology in case of high-speed motor with magnetic bearings and dry transformers
Weaknesses of VFD
– Lower power density
– Space in electric room (however, VFDs with enclosures for outdoor installation available as well)
– Inverter duty motor for some VFD models
Strengths of fluid coupling
– Simplicity of the solution
– Compact design
– High power density
– Low initial investment
– Shaving of torque peaks (transient torques)
– Suitable for outdoor installation and harsh environment
– Possible combination with planetary gear
Weaknesses of fluid coupling
– Lower efficiency compared to VFD, especially at lower speeds
– Large oil quantity –> environmental concerns
– High motor starting current (same as direct on-line start)
– Lower input power factor (~0.88 – 0.93) – power factor correction might be required
– Required space in machine room – difficult for existing installations
– Each driven machine needs its own fluid coupling
– Motor protection not included (only load protection)
– Very slow reaction time
– High torque at low speed is difficult / not possible
– More intensive maintenance compared to VFD
– Bypass of fluid coupling not possible –> thread for availability
– Redundancy on system level not possible
– Tailor made design –> difficulty to get technical details / guaranteed values in tender phase
Benchmark
Below you can find a comparison of fluid coupling and variable frequency drive using multiple criteria.
Fig. 4: Fluid coupling must always be inserted in the shaft string and requires precious space in machine room.
Fig. 5: Fluid coupling (with or without planetary gear) often requires a pony motor (most left component) to solve the starting issue [4].
Fig. 6: Comparison – Power factor and reactive power consumption of electric VSD (VFD) and hydraulic VSD (fluid coupling); VFD with diode rectifier considered (for detailed explanation refer to our previous post [5])
Fig. 7: One single VFD can soft start multiple motors and continuously drive one of them –> large space and cost savings
Don't believe everything in advertisement
Some manufacturers of fluid couplings have very aggressive marketing campaign (I tend to call it propaganda). They especially like to compare their solution with the VFD alternative and claim the fluid coupling to be better in every single aspect. These are some of the incorrect statements frequently used:
– Fluid coupling has higher efficiency than VFD
In a narrow controlled speed range the fluid coupling reaches fairly high efficiency. The losses in the hydrodynamic system are proportional to the slip, i.e. the relative speed difference of the impeller section and the turbine section. Therefore, at 98% of nominal speed the fluid coupling generates approx. 2% losses as part of speed conversion. When operating at 90% of nominal speed the losses in fluid coupling equal to approx. 10% of transmitted power in case of constant torque. This is physics – the larger the slip inside the hydrodynamic system the more losses and lower efficiency. Larger slip means that less motor power is transmitted to the output shaft and the difference are the losses heating up the fluid medium (usually oil). That is the reason why the fluid needs an external re-cooler.
Hint: You can easily check the cooling capacity of the oil re-cooler. If the fluid coupling has rating of 12 MW and efficiency of 97% then the losses make roughly 3% of transmitted power, i.e. some 360 kW. A good system has a reasonable design margin so the re-cooler would be rated 400 kW or 420 kW. However, if the re-cooler is rated 900 kW and still the 12 MW fluid coupling claims to have 97% efficiency then something is really fishy.
Some marketing material of particular manufacturers of hydrodynamic couplings show absolutely incorrect benchmark with VFDs: the fluid coupling has flat efficiency curve almost constant over wide speed range while VFD efficiency sharply drops at lower speed. Note that this is completely wrong and in fact it is exactly opposite. Efficiency of fluid coupling drops as the speed is reduced. This cannot be avoided as the losses are proportional to the slip. On the other hand VFD has very flat efficiency curve. The constant portion of losses is rather small. It is mainly the auxiliary consumption. By the way, we have completely neglected this auxiliary consumption for the fluid coupling, but of course the fluid coupling also has auxiliary consumption of the oil feed pump, control system etc.
Note that VFD also has the possibility to transfer the motor directly to the grid via synchronous bypass. In such way the VFD controls the speed of the motor when the process does not require full speed. When the demand increases and motor shall operate at full speed, the VFD can eventually synchronize it with the grid and get bypassed –> zero losses in VFD and input transformer at nominal speed. How could fluid coupling outperform here?
– Fluid coupling is much more reliable than VFD
Fluid coupling is hydraulic solution and therefore fairly robust. That is true. However, the system consists not just of the fluid coupling alone. There is the oil pump, oil filters etc. What is usually not much shown in the promotion videos is the control system of the fluid coupling. Also, mean time between failure (MTBF) is just one aspect of reliability. However, for availability the mean time to repair (MTTR) is equally important.
Still, to be fair, the MTBF of a fluid coupling is in the range 25 to 40 years and therefore higher than the MTBF of a VFD with 8 to 12 years. However, a repair of VFD usually consists of a replacement of faulty component. With maintenance friendly design, good access to internal components, trained personnel and available critical spare parts the repair takes just few hours resulting is very short mean time to repair (MTTR). The VFD’s fault logger helps to identify the failure very fast and efficient. And the rapid development in field of predictive maintenance and diagnostics open vast possibilities for further improvement. On the other hand, a mechanical failure of a fluid coupling is a severe event and the rectification takes few days in best case. Therefore yes, the probability of VFD failure is statistically higher, but the rectification of this failure is quick and impact on the plant availability is less. Failure of a fluid coupling has lower probability, but once it happens it means a significant downtime and rather high impact on production.
Finally, VFD might be equipped with a bypass that can be used as backup in case of severe failure. Fluid coupling obviously does not have such option and to remove it from the shaft line would take few days.
– VFD has significantly shorter lifetime
This is frequently abused argument of fluid coupling manufacturers. They argue that the fluid coupling is a mechanical solution and therefore inherently robust. VFD, in contrast, is electronic equipment and electronics as designed for short lifetime and meant to be replaced soon. In some absurd propaganda a VFD is compared to a personal computer or a smart phone intended to be replaced every 2-4 years. In similar absurd propaganda one could compare the fluid coupling to a conventional bicycle that is also a ‘purely mechanical device’. I use my bicycle daily and besides tyres I need to replace the chain every season. That is the official recommendation of both manufacturer and service shop.
Of course, above argumentation is an extreme simplification and it is not correct. The VFD solution is contact-less and there is no wear. The power semiconductors, specifically thyristor based devices, do not age. This is not a marketing, but a fact confirmed by the extensive field experience [6]. Yes, the control system is something that usually becomes obsolete as the first thing. However, also the fluid coupling has a control system (at least a primitive one). Or does anyone think that a remote compressor station as part of a gas pipeline is controlled manually by a personnel being locally present 24/7? Seriously? When we agree that every modern variable speed solution – no matter if electronic or mechanical/hydraulic – has a state of the art control system as its integral part then this point is neutral for all competing technologies. In fact, the VFD manufacturer is probably in a better position to perform a control upgrade as this is his core competence and usually his own product. In contrast, the mechanical manufacturer might outsource the control system and the continuity and availability of control upgrades is not taken for granted.
One famous manufacturer of fluid couplings claims that his spare parts are available for at least 30 years. Even more, it is claimed to be a guaranteed availability… What a courage! In today’s dynamic world who can be sure that company XY will exist in 5 years from now? Are such quotes firm promises or nothing more than pure marketing tricks?? On the other hand, some medium voltage VFDs are in operation for over 30 years. This is not a theoretical projection into the future but a fact based on field data.
– Fluid coupling is much simpler solution compared to VFD
Generally fluid coupling seems somewhat simpler – this is partly true. However, there are several additional components required along with fluid coupling: oil feed pump, oil heat exchanger (re-cooler), extra coupling to insert fluid coupling into the shaft string, oil filter, digital control panel, pony motor for starting (if direct on-line start not acceptable) and synchronization equipment along with voltage measurement. Overall, to be fair, it is a simpler solution, but it comes along with some limitations that VFD does not have.
Some manufacturers of fluid coupling, resp. combined fluid coupling + planetary gear solution push their product very hard and often make deformed comparison with VFD. Their frequent argument is the amount of components in VFD system. Note that e.g. harmonic filter is very seldom used or linked to particular VFD topology. Most VFDs do not require any harmonic filter. The rectifier technology (diode rectifier in combination with phase shifting transformer or active rectifier with harmonic elimination or minimization methods) fully complies to international standards such as IEEE 519, IEC 61000, G5/4 or others. On the other hand, fluid coupling might require additional power factor correction since power factor of asynchronous machine does not reach 0.95 as value requested by many utilities. So the argumentation can be easily reversed. It is also not true that VFDs require large and expensive air conditioning system. High power VFDs are almost always liquid cooled. Also the need for transformer is not a true argument. First, there are direct to line VFD topologies on the market that do not require an input transformer. Second, for high power applications the motor is connected to a higher voltage level (e.g. 33 kV or 110 kV) so a step-down transformer is required also with fluid coupling.
– VFD must be replaced every few years / spare parts are not available
Such statement is totally wrong. Instead of long counter-arguments you might ask some customers that use VFDs for decades. The LCI drives are being used since 1970s. The lifetime of VFD is at least 20 years. In fact, many customers require 25 or 30 years in their purchasing specifications and the VFD manufacturers obviously comply. Control upgrades are likely unavoidable during such long time period, but this is natural and nothing to be worried about. Renowned VFD manufacturers will always be able to provide control upgrades as part of their service capabilities and customer commitments. Availability of spare parts is as good as those for fluid coupling. The spares are available at least 10-15 years after the manufacturing of specific model is terminated. If sub-supplier of internal component is changed, backward compatibility is usually guaranteed. Last but not least – customer is always informed about the lifecycle status of his VFD.
Of course, there are always exceptions. If you buy a VFD from small company that enters bankruptcy short after delivery of your drive then you might end up with spare part issue. However, exactly the same could happen with manufacturer of fluid coupling if it is some ‘garage company’. As always, you shall compare apples with apples.
Summary
Both technologies, variable frequency drives and fluid couplings, are used in applications benefiting from speed control. Both of them help to save energy and therefore conserve the environment. Fluid coupling is very compact due to high power density and suitable for harsh environment. On the other hand it must be inserted in the mechanical lineup and requires space in machine room. It also does not solve the starting current issue and generally has lower power factor and much lower dynamics. Variable frequency drive might have slightly higher investment cost. However, this is fully compensated by superior efficiency, high power factor, excellent dynamic response, immunity against supply voltage dips or accurate speed control. Moreover, VFD entirely protects the electric motor without the need for external protection elements. VFD is extremely flexible and versatile – various cooling concepts, multi-drive arrangement, soft starter with synchronous bypass, solutions for outdoor installations (“NEMA 3R enclosure [7]”) etc.
In our view, not surprising, VFD is the better option providing all the flexibility required and having answers to all challenges around integration of variable speed drive systems. VFD offers lower life cycle cost and shorter payback time.
References
[1] Fluid coupling: Principles of operation (excellent video by US Army from 1950s – still actual), https://www.youtube.com/watch?v=tfe2ym8ojOw
[2] Energy efficiency – Part 5 (What efficiency figure to expect), http://mb-drive-services.com/energy-efficiency-part-5/
[3] How to choose a medium voltage VFD: Cooling type, http://mb-drive-services.com/choose-mv-vfd-cooling/
[4] Voith Vorecon start-up procedure, https://www.youtube.com/watch?v=NnvWYDbZKb4
[5] Energy efficiency – Part 6 (Power factor and reactive power consumption), http://mb-drive-services.com/energy-efficiency-part-6-power-factor-and-reactive-power-consumption/
[6] Experience with IGCT based VFD, https://mb-drive-services.com/experience-with-igct-based-vfd/
[7] NEMA 3R variable frequency drives for outdoor, https://www.youtube.com/watch?v=tfe2ym8ojOw
