by Jim Toepper, Moxa
Wind turbines are at the core of thought as they relate to wind power production.
Although this is true, other factors can help save tremendous amounts of money when installing or retrofitting wind farms.
The network plays a crucial role in the performance of a wind power system.
Using the appropriate infrastructure might seem easy at first, but without optimization, hundreds of thousands of dollars could be wasted, and, worse, you could be left with a system that does not perform optimally.
With all of the strict regulations, it is unacceptable to not have a system that is reporting with 99.999 percent up time.
Gone are the days of wind’s blowing on fan blades, which turns a generator and out comes electricity.
The amount of analytical equipment in the nacelle can be staggering.
In some of the most advanced installs, one might liken it to a mini data center atop a pole.
Now, that might be a slight exaggeration, but nevertheless, the amount of monitoring data that comes from a wind turbine has increased dramatically over the years, along with the size and efficiency.
These two things are related. With more analytics comes greater efficiency and the capability to upsize.
Logically, then, you can say that because of the amount of data available, many companies have been able to innovate and create better and more advanced wind turbine solutions.
But how is this data used? How does it get from the sensors, programmable logic controllers (PLCs) and embedded computers to a system that can aggregate and monitor it from a holistic perspective?
With these questions in mind, we may ask ourselves why the reliable transport of that data often is not thought of as a critical factor when coordinating a wind farm install.
During the past five years, many wind farm installs have used the least expensive, most basic equipment to connect each tower back to the monitoring station.
Most of these installs have been using unmanaged switches throughout their entire networks in the worst cases and in the best cases have taken an information technology (IT) approach to networking.
Both are inexpensive approaches to data connectivity, but both are akin to the aforementioned fan blades and generator scenario.
It’s not reliable enough, not manageable and outdated.
Can you imagine more than 300 GW of wind power worldwide whose power and efficiency could depend on an outdated, inefficient, unstable network?
Reliability also is a factor when dealing with Federal Energy Regulatory Commission (FERC) regulations.
Accountability for all power generated stresses the importance of a very high uptime network.
In some cases, noncompliance fines are so big that seconds of network downtime can cost many thousands of dollars.
Redundancy is a must to ensure this category of uptime.
Industrial networking technology has advanced along with innovations in the wind turbine industry.
There are several ways to achieve redundancy.
Spanning Tree Protocol
Spanning Tree Protocol (STP) as defined in the IEEE 802.1D standard is designed to eliminate loops in a network by cutting the network into a loop-free tree shape.
Networks are not allowed to have loops unless a special redundancy protocol allows it.
The network shown in Figure 1 is an example of a redundant Ethernet network that uses STP and includes a loop between switches 1 to 5.
STP uses an algorithm to find redundant links in a network and allow certain paths as backup paths to prevent looping.
When STP is running, packets sent from switch 5 to the root switch will go through switch 3.
When switch 3 disconnects or fails, STP automatically rearranges the connections by activating the backup paths to forward data.
Although STP resolves looping and achieves network redundancy, it has some drawbacks.
One is slow recovery time.
With STP, recovery usually takes 15 seconds after the spanning tree is established.
This recovery time is too long for wind farm applications.
Rapid Spanning Tree Protocol
Rapid Spanning Tree Protocol (RSTP) is much like STP, but to overcome the slow convergence (recovery) time of STP, the IEEE has released IEEE 802.1w and 802.1d-2004 standards to make improvements based on STP.
This shortened the recovery speed to about 1 second.
For a critical network such as one used on a wind farm, however, it is much better to be able to recover from a network failure in less than 100 milliseconds to ensure the reliability of the network.
STP and RSTP are open standards that many Ethernet switch manufacturers have implemented in their managed switch products, which makes it convenient for interoperability, but again, recovery time in a critical application is likely more important than being able to use a mixture of brands of networking hardware.
Proprietary Ring Redundancy
Ring redundancy is common in industrial Ethernet networks across many vendors.
It overcomes the recovery time problem of STP and RSTP.
Many ring redundancy technologies feature a guaranteed recovery time of a few milliseconds.
Ring redundancy ensures nonstop operation of networks with an extremely fast recovery time.
Depending on the amount of switches on the network and the manufacturer, you can recover from a network outage in the sub-20-millisecond range.
Take Turbo Ring from one manufacturer, for example.
If any segment of the network is disconnected or powered down, the network system will recover in less than 20 milliseconds by activating the backup path in a ring.
But there are still some drawbacks to ring redundancy.
Interoperability is a challenge because redundant ring technologies are proprietary to individual suppliers.
In addition, ring redundancy only supports limited amounts of rings in a single switch.
Given the variety of deployment sizes and shapes of a wind farm installation, rings, although they can achieve fast recovery, are not the optimal solution.
Chains and Open Rings
Chains and open rings are advanced redundancy technologies that solve the inflexible topologies of rings and the slow recovery times of STP and RSTP.
This technology also can recover from a network fault in less than 20 milliseconds.
By using this newer, innovative redundancy concept, network administrators can have more freedom in the shape of the redundant topology among any network segment.
Chains and open rings support the perfect topology shapes for many new wind farms.
Chains and open rings work by connecting several Ethernet switches together to form a daisy chain where each side of the network segment is connected to the rest of the network without requiring routers.
With the chain and open ring methodology, network engineers can create as many redundant connections as they have switches and ports by linking a new chain or open ring to any segment of the Ethernet network.
Some vendors even allow for the addition of chains or rings to happen without any reconfiguration of the network.
This means that the addition of wind turbines–network expansion–is easier than it ever has been and reduces costs while optimizing the network.
The best part about this topology is that chains and open rings work with existing network architectures.
If part of your network is IT-oriented and uses RSTP or is unable to be changed, you still can connect a chain or open ring to it; they are compatible.
This would mean that at least part of your network would have the fastest redundancy scheme, and the other older part would continue to function as originally intended.
Using the latest technology provides better redundancy, flexible architecture and compatibility with IT systems, and it also saves substantial money.
Because of the unique chain and open ring topology, one fewer connection and two fewer ports are required.
The saved connections are those that are typically long-distance runs.
For instance, for RSTP and rings, your network diagram in the best-case scenario would look like Figure 4.
The red dashed lines are required redundant connections for ring coupling.
As you can see, by using a chain or open ring architecture, bandwidth is optimized, redundancy is fast, configuration is easy and you need fewer connections.
If this approach is taken while planning the next wind farm, many miles of fiber-optic cable as well as the trenching required for installation can be saved.
Add these two costs together with the cost of the saved fiber connections on the switch, and, in most cases, the newer more advanced network will pay for itself many times over compared with taking a more traditional connectivity route.
Although a lot of the aforementioned might seem complicated or new, it is accomplished easily.
Best, some have seen savings in the hundreds of thousands of dollars.
This is a newer, more advanced approach for executing the connectivity required in wind farms, but just as the wind turbines have evolved in efficiency, reliability and cost, so have the networks that connect them.
Many project managers and network engineers will be happy to see a way to decrease the total cost of an installation.
What’s even better, these cost savings can be achieved solely by considering a different communications approach.
This concept has been around for a few years and is one of those best-kept secrets, but communications requirements have increased so much over the years with wind farms that it’s time to look at a new way to do things.
With so much money to be saved and so much reliability to be gained, this newer way of thinking is likely something most will not want to overlook when it comes time to build a reliable communications infrastructure.
Jim Toepper is business development manager of power at Moxa. Reach him at email@example.com.
PowerGrid International Articles Archives
View Power Generation Articles on PennEnergy.com