Welcome Because the field of grounding is so broad, todays presentation is intended not to give you the keys to the gold mine, but to give you a handful of nuggets to you to invest in. 1
Ladies and gentlemen, we will be crossing the floors here between the electrical and instrumentation installations. The reason for this is that the majority of noise in one division, is often caused by the practises, of the other. Grounding and noise is a very dry subject hence the short practical break to let our minds recover, as I whisk you on a rapid and broad tour of the electrical spectrum 2
Personnel and plant safety Safeguard against conductor voltage stress Path for currents allows installation and operation of protective devices Minimizes overheating 3
Power grounding (Functional) or earthing or bonding. The general principle is the EEBAD, earthed equipotential bonding and automatic disconnection. The aim is to produce a low impedance path for fault currents to pass through, so that fault currents may rise high enough and quick enough for protective devices such as breakers or fuse to operate- and reduce the risk of fire. One loose terminal, or corroded lug can increase the impedance, to a point where ignition of surrounding structures can occur. 4
Need I say more, grounding in its self is primarily there to protect against fires from electrical equipment or cables. Life safety is provided by GFCIs. I have personally seen Motor control centres, starters, and panel boards where faults have not cleared in a reasonable time, and have caused fires. Try not to laugh at the next slide 5
Note the lack of understanding shown here Unfortunately this isn t a faked photo. 6
This is a normal fault loop path. 7
What we have here is a very good example graphics of the Instrument or DC ground for a installation We have 2 grounds one insulated from the enclosures one not. We are creating 2 paths of current flow- one clean for instruments, one general purpose. Technically they are all bonds other than the main connection, but most people still call them grounds. Notice they both still go back to the same service connection. 8
High frequency currents mainly use the skin effect to pass through a conductor, hence the common use of braids for ground bonding for EMC compatible enclosures, VFD s and radio signals. A high frequency in control terms is recognised as being >10kz, this would most often be derived from electronics, both low power and high power. 9
Most of you might have seen the installation of a VFD on standard Teck cable. With the Teck bonded directly to the structure of the MCC. What most of you might not have realised is that this is producing noise into the local area, both conducted, and emitted. An installation I was tasked with commissioning had a very small VFD, 1/3 of a hp, it was considered too small to need proper mitigation practices. We soon found out that every time that dosing screw energised, the high level ultrasonic level switch in the hopper bin switched on- causing much consternation to all those commissioning, to the point where the electricians had a moment of levity, and put aluminium foil over their hardhats, just like the film, signs. People normally expect this vfd noise to occur at much higher levels of power- to the point where in Europe there are regulations concerning the installation of any VFD, greater than 7.5kW, which then require noise, and harmonic mitigation from the builder of the drive. The easiest installation method is actually to use cable systems that shield the noise- similar to an shielded analogue cable, hence the use of Drive RX cable amongst others, which include copper shielding in the construction. 10
However this cable must be properly installed, it should pass through the mcc shell and be grounded directly on to the chassis of the drive, and not connected to the motor end- to avoid a ground loop. Functional safety bonding of the motor should be by a separate external bonding conductor. It should not be within the cable. 10
Some examples of Shielded VFD cable, there are others. 11
Manufacturers own technical literature recommend this practice, but its rarely followed. I have found ABB drive literature to be the best shown examples of the practice. I have found their methods do work. Filters should be applied to the input and output of the drive, to reduce the complex waveform which travels both upstream and down stream from the VFD. Insulated Bearing should be specified at initial selection, only a small adder then. Other methods of bonding only provide marginal increments of noise reduction. Following this all this avoids a major root cause of instrumentation noise, and hence increased project costs caused by delays. 12
Another reason its important, is that the complex wave form of noise, can interfere, with the switching of signals from powerline carrier communications devices. 12
This is the electrode formed around the structure, that is designed to intercept the lightning strike and divert it to the earth, to minimise physical damage from the strike. 13
What we have to remember is that a Lightning strike passes like a massive wave over the electrical network. Really its like a tidal wave, with combined effects which are all hazards in them selves. At each place on the network the potential will be different, and even around the network in the physical structure there will be voltage present. This is where a number of different hazards all come into play at the same time, with the following Touch potential-the obvious potential risk. Step potential- where the position of your feet and and resistance of your footwear can have a potential difference, which produces enough voltage to 14
be hazardous. Farmers are the aware of the impacts of the step potential, with livestock being particularly sensitive to the effect. 14
This is where critical equipment is protected from the voltage surge in the region of the lightning strike. Often by means of isolation transformers, filters, surge protectors.. Four our purposes tonight, one of the big ones we have to consider is communication equipment. Particularly the effect on radio, and telephone systems. One thing we must consider is that lightning is made up of 2 conditions- ionisation, and the actual discharge- the strike. There are many examples of radio equipment being damaged, as the combination of antenna, and high frequency signals, predispose the ionisation of the localized atmosphere, which triggers the lighting strike. 15
Lightning protection appears to cause problems for a number of people. Its purpose is to divert the energy of the lightning strike, around the structure via a low impedance path, to the general mass of ground. There are many areas where a lack of understanding of the elements can cause issues. One site I recently came across, was a first responders repeater radio, which consisted of 3 distinct element, the main antenna line of site with base station, the main target for any lightning strike. The driver electronics, The heart of the system, with frequency hopping high frequency components. And the repeater antenna. This repeater was longditudinal down a 620m tunnel Here the issue was that the electricians confused the grounding for the electronics element, with the grounding for the main antenna. They were complaining about needing to get the 2/0 ground from the grid down the 35mm conduit feed the power to the drivers, which only needed a 16
#6, without looking at the drawings properly, which indicated the 2/0, which connected the antenna to the ground grid directly, without passing through the control building. 16
Surge arresters There are many to choose from, these form are often formed from banks of Metal Oxide Varistors, forming the equivalent of a diode, which breaks down under high voltage, and diver t the dangerous voltages away from the equipment. These should be applied to signals where they enter and leave building. To minimise strike impacts. Be aware through that surge arresters do have a finite life, of stress events, which should be checked. Depending on the level of integrity required the status of these devices can be monitored, by the control system. Personally I recommend maximising the use of di electric free or metal free fibre, for communications, as this avoids the risk of interference. 17
Wake up 18
I would like to assume the person is holding a test instrument and is happy with the result, but I guess they may be praying to the electrical gods. I offer a beer to the first person, who can find the word TEST in section 10-Grounding of the CEC. It s a safe bet- its not there. 19
NFPA & IEEE recommend 5 ohms or less. Telecoms 5 ohms or less. Fall of potential requires disconnection. Selective measurements no disconnection. 20
Loop test is performed live. Measures resistance of the total supply and ground. Through transformer, supply cables, grounding, transformer XO connection. 4 sites measured; using FLUKE 1651B. 3 sites from 4 have grounding / bonding problems. 21
The standard 4-20mA analogue signal is considered fairly noise free, provided a few simple rules are followed, only bonding the shield at one end, so that circulating noise currents are avoided. Also keep away from high power elements such as motor cables- lets refer back to the VFD cables. Co axial cable can be used, however Its expensive. Its also not very common- more an aircraft use. A very common solution to noise is the use of twisted pair cabling, which reacts like a balanced spinning top, and rejects imposed noise, returning to its happy place all the time. 22
Here we have a Hart signal super imposed over an analogue signal The Hart signal is the main item sensitive to noise, as it s a modulated waveform imposed on the carrier analogue signal. 23
The main reason for low impedance bonding is that the differences between terminations can create a voltage difference on the cable, which in turn creates circulating currents, which can add or subtract from the sample signals values. This is of particular concern in the analogue to digital conversion circuits. The image on the left is an analogue instrument that s been grounded at both ends of the shield. The image on the right is an alternative to just one grounded connection on the shield, is to isolate it, models of isolator can also be used to split the signal for redistribution to telemetry systems say. 24
These are instruments that we find every where. 25
Magflo style flowmeters are sensitive to noise, in a couple of ways. They insert an RF into the flow, and measure the signal downstream. On many installs they require grounding rings, such as those on the right, to provide a noise free location, in which to provide the RF signal. Refer to manufacturers info for their recommendations, particularly if there is cathodic protection, as isolating power supply s may be required. 26
The common fieldbuses such as Modbus RTU, Foundation Fieldbus, Device Net, and Profibus, all work from common reference voltages, hence the best practice installations have separate bonding conductors as well to provide reference. This is due to the comparatively high speed of these networks, and to maintain the required square wave forms. The most common is to make sure that the shields are connected together, then bonded at one place only. 27
Here we show a selection of the different cable types for some of the 385 odd industrial fieldbuses. Clockwise from the top right we have --Foundation Fieldbus. Profibus, Ethernet, DeviceNet. DC Reference (or instrument )grounding A low impedance ground generally in the region of 1-3ohm between the ground grid, and the ground rails. It is connected to the ground grid (there is by code only one ground point- that s at the source of the supply, all other points are deemed to be bonds). This is often a number of insulated bonding rails within a control panel to which all the signal and communication cables are bonded. Best practice is to use Fibre Optics where possible- although with FF cables this cant be done. Most common use of fibre is IT. 28
Diagnosis of fault on these networks does require more technical knowledge, and special equipment. One of the best tools here is an oscilloscope. As can be seen from the pictures, a certain amount of noise can be seen. There are often surge arresters built in to the communication adapter cards of instruments, it has been known that these can cause network communication failures 29
DCS manufacturers such as Emersons Delta V have a very good resource. Shield your noisy cables, and filter the harmonics, filter with signal isolators. Use fibre on networks where possible- also aids future protocol migration. Connect the grounds where required. Single end to avoid ground loops. Test it. Test it, and Test it. 30
31
32