|
Surge Protection - LAN 1. Introduction This article explains how to apply surge protection to LANs and helps you understand the following questions: u What is surge protection? u Is our LAN at risk and should we have surge protection? u What are the costs and benefits of surge protection facilities? When u have surge protection, how do you apply it to your LAN? u Which surge protection product do you choose? Surge protection is the core business area of Surge Free Co., Ltd. This article is not a thesis on networks, and since the topic of networks is so vast, it is assumed that the reader has basic knowledge of networks and electricity. Most people are familiar with networks, but are just starting to learn about surge protection. Even though they know that surge protection is necessary in fields other than networks, they are unable to apply specific measures. 2. What is surge protection ? Any device that uses electricity can be damaged or completely destroyed by a surge at any time. A surge is a voltage that is much higher than the normal voltage and appears in a system such as a LAN for a very short period of time. It is also called transient over-voltage, spike, impulse, or noise. Surges can also occur when large loads are switched on or off nearby, or when large currents are interrupted due to power system faults, but the most powerful surge is lightning. In particular, direct lightning strikes to buildings can cause catastrophic damage. can cause damage to electronic components. Thankfully, however, this is extremely rare. Typically, damage to electronic components is caused by lightning strikes that can occur several kilometers away. Typical damage to PCBs includes bulging or blackening of copper plates, burnt-out resistors, and flying IC chips. In addition, surges that appear on data/communication lines can sometimes cause errors in program execution. Typical examples of program errors caused by surges on data/communication lines include programs stopping mid-execution or the keyboard becoming locked. Although less severe, what is very disturbing is the potential damage that surges can cause to semiconductors. This damage can appear months later and appear as if it were caused by static electricity. Surge protection uses a hardware called SPD, and when this device is properly installed, it reduces the surge voltage to a voltage that is safe for the equipment. The operation of SPD will be discussed separately later. SPD: Surge Protection Devices 2.1 How big is the surge ? Lightning-induced surges can be defined in terms of voltage, current, and time. A. Open-circuit voltage: The maximum voltage that can be applied between cables without breaking down the cable insulation. B. Current (short-circuit current): The maximum current that flows when the cable insulation is destroyed. C. Time: The time it takes for a voltage or current pulse to reach its maximum and then decline. The rise time of a lightning impulse is several μsec, but the decline period is relatively long, lasting tens to hundreds of μsec. |
||
Research has shown that surge protection devices must be able to handle voltages ranging from several kilovolts to tens of kilovolts and currents ranging from several kiloamperes to tens of kiloamperes.
2.2 Surge Protection and Network Layer
(Net-work Layer)
Defining all of the networks would be a huge undertaking, so we will talk about the well-known OSI 7 layer model (Figure 1).
Figure 1. OSI 7-layer model
SPD is installed in the network and also installed in signal lines including telephone lines and power lines. Basically, SPD is installed in the signal lines of the physical layer, and SPD is also installed in the data link layer, which is the second layer. For example, SPD should be installed in the physical layer modem, telephone line, LAN UTP cable, etc., and also installed in the second layer HUB, LAN card, router input/output, etc.
Here are some things to consider when investing in surge protection. The decision to implement surge protection should be based on an economic assessment that takes into account the likelihood of damage to the system and the impact on the surrounding area.
As an extreme example, consider a petrochemical plant that handles flammable or explosive liquids or gases in an area where lightning is frequent. If the remote metering or control equipment fails, a huge disaster can occur.
Another example is a TV that breaks down in an area where lightning rarely occurs. This may happen once in a lifetime, so you can just buy a new TV.
3. RISK FACTORS
As with insurance, the only thing that is certain about surge protection is its cost. Risks can only be expressed in terms of probability. The same place may not have suffered damage from lightning for 20 years, or it may have had two in the space of a week.
British Standard BS 6651: 1999, Appendix C classifies hazards as follows:
u lightning frequency
u Size and exposure of the building
u Resistance of the earth (the higher the resistance, the higher the risk)
u The number and length of copper cables entering the building (including power lines, telephone lines, and data lines)
u Vulnerability of equipment
|
4. Economic factors Even if the risk factors for equipment damage are sufficient to introduce surge protection, there are still obstacles to consider. Not only do quality SPDs cost tens of thousands to hundreds of thousands of won, but their installation costs are also high. In recent years, the prices of computer equipment and network equipment have dropped significantly. Therefore, the cost required to establish surge protection is an issue, and especially in small networks with only a few computers, replacing a damaged network card or PC is cheaper than the cost of surge protection. However, you should keep the following in mind: Restoring a network is not simply a matter of quickly replacing a single piece of hardware. u There may not be a technician who is familiar with the on-site situation. u Finding broken H/W is not easy. u When multiple pieces of H/W fail, it becomes more difficult to find the failure. Replacing H/W is not always possible. There are H/Ws that are considered completely useless for normal business purposes but need to be replaced very quickly. Is it possible to recover lost or discarded data? Insurance companies in some advanced countries are advising that power and communication systems be equipped with surge protection, and are applying this as a premium factor. In fact, in the United States, building completion inspections are conducted by insurance companies, not government agencies. Zurich Insurance is offering a 10% discount on insurance premiums if surge protection is provided with HAKEL SPD, which we handle. H/W can be easily replaced with insurance. However, data is lost. Losing data means interruption of business processes, and the resulting damage is more than the calculated amount. 5. How do surges threaten networks ? This section covers: ① The biggest threat to networks is the surge between ground and ground caused by lightning strikes nearby. ② Direct lightning strike to the building surrounding the LAN A relatively small but fatal threat to the network ③ Intermediate threat to LANs with intermediate number of lightning strikes on power lines, power line accidents, and power line surges due to switching Ground surge caused by lightning strike nearby (Ground Potential Surge) A single lightning strike contains several strokes of different sizes. During the discharge of electrostatic energy in a thundercloud, currents of several tens of kA flow through rocks and soil on the ground surface. Lightning currents of 100 kA and 200 kA are often used to describe the worst direct strike. Since the earth is not a perfect conductor of electricity, when lightning current flows, extremely high voltages are induced between various points on the earth's surface. For example, if a current of 100 kA flows through a ground resistance of 10 ohms, which is not that high, a voltage of 1,000 kV is induced between several meters or tens of meters from the earth. This is a ground surge. If two buildings are located in a lightning zone and are separated by some distance, a potential difference will occur between the two buildings. The electrical grounds of each building, which are connected to the local ground through rods and mats, will experience a large potential difference between them. If the electrical system of each building is completely contained within its own building, this potential difference will not cause any problems. However, if a cable such as a LAN connects two buildings, the equipment in one building will experience a difference in ground potential relative to the equipment in the other building, and this difference in potential is sufficient to destroy the insulation of the LAN equipment. Figure 2 illustrates this. Today's buildings are inevitably interconnected by various cables for electricity and communication. However, if the cables are interconnecting buildings, considerable care must be taken to install SPDs. In fact, cables outside the building are completely exposed to transients. The building contains numerous signal cables that are not connected to the building's electrical ground. |
