Whether a distributed or centralized architecture is right for you, the choice is yours with the VibroSmart and VM600Mk2 platforms from vibro-meter.
The debate over centralized versus distributed instrumentation architectures has been ongoing for more than 40 years, predating many of those reading this. Nor will it end any time soon. This is unfortunate because it is needless.
Rather than viewing these architectures as one versus the other it is helpful to view them as one plus the other. Both have their place. Both make sense when the application particulars dictate one approach in one situation and the other approach in another situation. For many manufacturers, however, their distributed offerings entail very different functionality and channel types than their centralized architectures, forcing customers to choose based not on the best architectural fit but the best feature set.
We have a refreshingly different perspective: give our customers a similar set of features and channels in each platform and let them choose based on the architecture that fits the application best – not the feature set. Nowhere is this more evident than in our hydro monitoring capabilities. For applications that can best be addressed by a distributed architecture, we’re pleased to offer our VibroSmart platform. And for applications that can best be addressed by a centralized architecture, we’re pleased to offer our VM600Mk2 platform.
If you have never actually installed a monitoring system (or had to pay for it), it is easy to think that the preference for one of the two architectures might boil down simply to subjective factors. But in reality, it primarily boils down to something quite objective: installation costs.
Consider this important fact: the total installed cost of wiring typically runs about €16 per meter per point and is often the single most costly part of a monitoring project. To make this more tangible, let’s consider two scenarios. The first is three machines with 8 monitored points each, located 250m from a control room but for which wiring already exists and with an existing monitoring system that is obsolete and ready for replacement. The second scenario assumes the same three machines but will use a distributed system. This is depicted in Figure 1.
“The total installed cost of wiring typically runs about €16 per meter per point and is often the single most costly part of a monitoring project.”
Figure 1: Centralized versus distributed architectures can result in dramatically different field wiring costs for some applications. In this instance, the wiring costs for a centralized approach are 87% more than those for a distributed approach.
If we had to install the wiring in scenario #1 from scratch using today’s prices, the cost would be about €96,000. Fortunately, the wire is already in place and can be reused. In this case, we might simply choose a solution that can be mostly easily mounted in place of the outgoing system – also in a 19” EAI rack-mounted form factor – and the VM600Mk2 with its centralized architecture would be a good fit.
In scenario #2, selection of a centralized architecture has a very different outcome. As in scenario #1, the wiring costs would be about €96,000 if the wiring for each and every sensor went to a system located in the control room. However, by use of a distributed monitoring system we now have an option that allows us to keep individual sensor wiring for each system to no more than 10m (assuming we mount the distributed systems at each hydro unit) and then to run redundant network cables back to the control room where a human-machine interface (HMI) will be located. In this instance, we now have two cables that run 280m (250 + 15 +15) and 24 cables that run 10m each. Total wiring cost using a distributed architecture becomes €12,800. Or in other words, a savings of 87%! This is substantial and underscores why a customer would have a compelling reason in some applications to choose a distributed versus centralized architecture. It is recognized that each distributed monitoring system must generally be mounted in its own simple junction box, but this will rarely eclipse the savings enabled by shorter wiring runs and becomes more acute for longer runs.
Above, we showed how a distributed architecture could dramatically reduce wiring costs in a scenario with three machines. This can also hold true for a single machine where the VibroSmart system is distributed around the machine to keep wiring runs very short. The VibroSmart architecture incorporates a virtual backplane that can be created both by plugging modules side-by-side using special connectors on the terminal bases, and/or by running Ethernet cables between the modules. This is called the “sidebus” (or S-bus) in the VibroSmart system and allows adjacent modules to be connected side-to-side using the special S-bus connectors and non-adjacent modules to be connected via conventional Ethernet cabling.
This granularity of distribution can be especially advantageous in applications such as hydro where vertical machines may span multiple floors of the power hall and a small junction box with one or more VibroSmart modules can be mounted very close to each cluster of desired sensors. As shown in Figure 2, this can be a very small junction box to accommodate just one or two VibroSmart modules; the power can be supplied remotely, meaning the junction box does not even need to hold individual power supplies. A single Ethernet cable can be run between each junction box to create this virtual backplane (S-bus) and a single twisted pair can carry the required 24Vdc power needed for each junction box.
Figure 2: The configuration shown here was provided for a hydro customer in eastern Canada, allowing consolidation of two separate systems (one for air gap and one for vibration) into the integrated environment provided by the VibroSmart platform. Modules were distributed around the machine to minimize wiring runs and keep installation costs correspondingly low.
As shown in Figure 3, Hydro machines have a number of special measurements such as air gap, magnetic flux, partial discharge, and cavitation – most entailing special sensors and corresponding special signal conditioning. They also have a number of conventional measurements such as temperatures, flows, thrust position, phase, casing vibration, and shaft radial vibration; however, there are special signal conditioning requirements for vibration on hydro units that are not characteristic of other machines.
One of these characteristics is a requirement for very low frequencies. Smaller units such as those incorporating Pelton wheels will often turn at several hundred rpm. While considered quite fast for hydro units, this would normally correspond to a very slow machine in and of itself – much slower than most motors, pumps, compressors, blowers, and gas and steam turbines. Even so, larger units using Francis and Kaplan type designs typically turn even slower and speeds of only 50-60 rpm are common. This means the sensors and monitors must have a frequency response down to 1 Hz just to capture rotational speed, let alone sub-harmonic frequencies. For which fractional frequencies down to 0.1 Hz would be required.
Figure 3: Hydro turbine-generator machines have a number of conventional as well as specialized measurement requirements. Even on conventional measurements, such as radial vibration, requirements are different than on other machine types such as the need to address very low frequencies (down to 0.1 Hz) and the ability to provide signal conditioning such as NOT 1X that can detect the presence of rough load zone.
Another characteristic behavior of hydro units is so-called “rough load zone” (RLZ) that corresponds to turbulent flow conditions when bringing a unit online and adjusting flow via the wicket gates. It is typically detected from the vibration signature by removing the 1X component from the overall broadband signal, resulting in a measurement known as NOT 1X. Operators required rapid feedback from a monitoring system when a machine is in RLZ because it can inflict significant damage if left in the condition for too long.
Some hydro units are part of pumped storage schemes whereby power is generated during periods of peak demand by allowing it to flow out of a reservoir, spinning the turbine and connected generator normally. However, during periods of low demand, this water is pumped back up into the reservoir by running the generator as a motor and the turbine as a pump. This allows the hydro plant to act as a “battery” by charging and discharging. Charging the battery occurs by pumping, storing the power as potential energy by virtue of elevation of the water in the reservoir. The battery can then be discharged when required by allowing the water to flow back through the turbine, spinning the generator and producing power. This cycle usually continues daily to charge during off times and discharge during peak demand times. Such plants require special monitoring schemes that can detect when the machine is running in generating mode versus pump mode because the alarm setpoints will be different. This is so-called “state-based” monitoring that recognizes the operating state the machine and adjusts the setpoints accordingly.
One of the substantial advantages of the vibro-meter approach is that these hydro measurements – and many others for other machine types – can be made in both the VM600Mk2 and the VibroSmart platforms. This ensures that you can choose your platform based on what architecture works best for your application – not because the necessary functionality is available in one platform but not the other. This is enabled by vibro-meter’s “one card does it all” approach to signal processing. When we designed the original VM600 back in the late 1990s and launched it to the market in 2000, we made a ground-breaking decision to create a truly universal monitoring card that could make any vibration, position, or speed measurement – and indeed any dynamic or quasi-static measurement – in a single module. This was known as the MPC4 (Machine Protection Card – 4-channel) and it revolutionized the industry. That same approach was used in the design of the VibroSmart platform where a single module type (VSV30X) could be used universally for any conventional dynamic or static machinery measurement – whether vibration, air gap, cavitation, magnetic flux, etc. This is done by allowing the user to fully customize the transducer type and the measurement particulars.
Figure 4: The universal “one card does it all” approach of the MPC4 in the VM600 platform was used as the basis for the VibroSmart platform as well. The VSV30X module is a universal module that provides all measurements with two universal dynamic channels and one speed/auxiliary channel.
It is this approach that allows both the VM600 and the VibroSmart systems to be used across virtually any machine type for any necessary sensor input. In addition, both platforms now combine the condition monitoring functionality and machinery protection functionality in the same module/card, reducing the burden of spare parts even further with a truly universal approach. They also incorporate integral relays, eliminating the need for separate modules to provide discrete outputs.
A defining characteristic of vibro-meter’s monitoring architectures is simplicity. This is evident not only in the “one card does it all” universal approach to channel types, but also in the relatively small number of other modules required.
Figure 5: The simplicity of the VibroSmart system is evident in a system that requires only two module types: one for communications (shown here) and one for universal signal processing applications including the suite of measurements used for hydro machinery.
For the VM600 platform, there are only five other modules: temperature, power, CPUM (communications), RLC (additional relays), and XMV/XMCi(condition monitoring).
In the VibroSmart platform, it is even simpler and consists of a single additional module type: the VSI010. This module supports both serial and Ethernet-based communications with automation and control platforms using protocols such as Modbus, Profibus, and GOOSE. This allows compatibility with older machinery control and process control environments using serial communications as well as newer environments with Ethernet-based communications infrastructures. The module also supports full network redundancy to ensure that communications are never interrupted.
Accessories are available such as network switches and BNC patch panels.
The agnostic approach taken by vibro-meter in its hardware platforms also extends to the software connectivity. Both platforms are fully supported in our VibroSight suite supporting configuration, condition monitoring, decision support, data import/export, and more. This ensures that a single software environment can unify all of the monitoring hardware you use across your operations, allowing you to mix and match hardware based on the particulars of your operations, machine locations, and new versus retrofit installations. VibroSight also contains the specialized plot types used on hydro machines such as air gap and magnetic flux. This ensures a comprehensive solution for hydro machinery that incorporates all measurement types – not just conventional vibration and position.
Figure 6: The VibroSmart and VM600 platforms are fully supported by VibroSight software including the diagnostic rule box “decision support” capabilities (top) and the specialized plot types such as air gap (bottom) unique to hydro machines.
Our ability to address the hydropower market extends beyond just the flexibility of both centralized and distributed monitoring systems, and corresponding software with our VibroSight suite. We also offer the specialized sensors that are routinely employed on hydro turbines and generators, along with conventional acceleration and proximity sensors. You can learn more about both the VibroSmart and VM600Mk2 platforms in our online product catalogue or by contacting your nearest vibro-meter sales and service professional.
Figure 8: In addition to monitoring systems, vibro-meter provides a complete complement of sensors for the aggressive environments, low frequencies, and specialized measurements encountered on machinery in the hydroelectric industry.
i XMV/XMC modules are 16-channel devices and required only when condition monitoring is required and machinery protective functions are not needed. When machinery protection is required, the MPC4Mk2 modules are used and provide integrated protection and condition monitoring, eliminating the need for a separate module such as the XMV.