Learnt a great deal from Amir Boshi of Satec and Powerqualitydoctor.com earlier today.
Thanks to Adrian Teo of Efunity for the kind invitation.
Learnt a great deal from Amir Boshi of Satec and Powerqualitydoctor.com earlier today.
Thanks to Adrian Teo of Efunity for the kind invitation.
To many of you in Singapore, you will only probably observe lights flickering either when 1) the light bulb itself is due for replacement or 2) during the very brief moment where there was an electrical fault in your area.
You will also not find the word ‘flicker’ in Singapore’s Transmission Code. but that does not mean there are no flicker limits imposed for Singapore’s electrical network. In Section F2.1, it states that “…………shall be in accordance with the requirements set out in Engineering Recommendation P28 of UK.” Based on this old ER P28, the limits are 1.0 and 0.8 for short (PST) and long term (PLT) severity values respectively.
So what is flicker, you may ask?
Flicker is a power quality problem primarily concerning human’s perception of changes to the output of the light bulb. These changes are caused by voltage fluctuation due to electrical loads with rapid variations in its loadings. An arc furnace is often cited as an example.
PST 1.0 refers to the level of voltage fluctuation that will cause more than 50% of you to notice and complain. It is based on the changes of the light output of a 60W incandescent light bulb. One can measure these flicker values either using a Flicker meter or a modern power quality analyzer.
So are there flicker issues (exceeding limits) in Singapore? I will say, there are flicker issues here but not necessarily a problem. Remember, the limits set upon was based on the incandescent light bulb, which you hardly come across today.
An interesting study (Cigre 449) conducted by Cigre Working Group C4.108 revealed in a limited test of their own that modern lighting are less sensitive as compared to the 60W incandescent lamp, when tested under instantaneous flicker value of 1.0.
So there could very well be flicker issues in your area. It’s just that modern lighting have saved you from being irritated.
Below are some of flicker trend values in Singapore from my past work. Yes, flicker do exists here.
I had the privilege in toying around with the new HDPQ Xplorer earlier today, thanks to Terry Chandler of Power Quality Thailand. Definitely re-affirmed my belief that Dranetz PQ instruments and Dranview (especially) are at least one notch higher against its competitors.
p/s: Thanks Terry for visiting us today.
Mr. Harmonics is frequently being blamed when an equipment failed (or when a cable burnt, capacitor bank blown, or a circuit breaker tripped without an obvious fault). Some without due consideration of other simpler factors will blame Mr. Harmonics and his cousins like Mr. Resonance (or perhaps his friend, Mr. Transient – story for another day) straightaway.
Surprisingly, it is a fairly easily accepted reason here. And with power quality instruments getting more affordable these days, it has been becoming quite common to see someone using this new toy, measure current harmonics in percentages of 50-80%, and straightaway concluded that it is indeed a harmonics problem.
Firstly, when it comes to harmonics, we need to know; is it voltage or current harmonics? While there is a common indicator to measure both of them – using the Total Harmonic Distortion (THD) formula (RMS value of the harmonic content expressed as a percentage of the fundamental), one needs to know the pros and cons of using such indicator when applying to voltage/current harmonics.
Usefulness of THD
Limitations of THD
Here in Fig1, is the current waveform and spectrum of the common switched-mode-power supply to our PC/laptop at work or home. Looking at just its THD% current, one will be extremely alarmed (162%!!). So should all of us purchase harmonic filters for all our homes/offices then? (fact: the actual amperes of this circuit is less than 0.6Amps, and VTHD is only 1.52%)
When it comes to current harmonics, it will be more meaningful to use other alternative indicators such as Total Demand Distortion (TDD), or use absolute amperes (my personal recommendation).
Fig 2 and 3 shows the trending results of current harmonics, presented in THD%, Harmonic Amps and TDD%.
Total Demand Distortion
It was reported that the voltage disturbance (lasting about 100ms) yesterday at about 1418hrs was caused by lightning affecting the inter-connector link between Singapore and Johor, Malaysia. Link from the statement by SP PowerGrid.
(FYI: Singapore is inter-connected to Johor at 230kV – a transmission level voltage; any disturbance at this level will typically be ‘seen’ by everyone in Singapore with 1/4 of the island ‘seeing’ the worst dip values).
While there is hardly any blackout moments in Singapore due to how the grid is designed and operated here, voltage dips like the above can actually be quite common. In the last 6 months, there has been 3 of such transmission level-originated voltage disturbances (as taken from our monitoring records; 5/11, 10/9 and 27/7). As electrical faults in the grid (be it caused by cable damage, customer’s installation fault, utility equipment fault or lightning) cannot be totally eliminated (only reduced thru rigorous maintenance, regulations/enforcement etc), customers with sensitive equipment should take measures in protecting their equipment adequately. They should also be aware of this Clause in the Transmission Code (which is also in their Connection Agreement with the Utility).
This clause basically rules out any ‘compensation claim’ from the Utility due to losses sustained arising from voltage dips. To some, this may seem ‘cruel’, but I believed this clause is fair as the electrical network is interconnected and everyone (Customers, Utility, Gencos) has to play their part. From my experience, the awareness on the need and know-how for protection against voltage dip is still very much work-in-progress here in Singapore. The key to solve voltage-dip related problems is not to seek compensations, but rather to undertake a proper assessment on the vulnerabilities of their equipment against voltage dips and then invest in the right mitigation solutions / methods.
“Explaining the outage, SGX said power is supplied to the data centre from two separate substations, which is then connected to the individual UPS systems. A “momentary fluctuation in power supply from the substations” caused the UPS systems to switch to its internal power source, but these power sources malfunctioned.” – ChannelNewsAsia.
In the early afternoon today, there was a voltage dip recorded at approximately 1418hrs (2:18 PM). A snapshot of this dip waveform captured in one of our sites in Bedok area is shown below. We also received several calls from various sites scattered over the island, reporting chiller operations being affected, etc. These could only mean one thing; a transmission level (eg. 230kV) fault had occurred.
Coincidentally, a report from ChannelnewsAsia reported an incident at SGX at approximately the same time, as shown in the following news snapshot. Without knowing further detailed information, I cannot really confirm / comment if these two incidents are related.
Voltage dip, while it typically occurs for just a slight fraction of a second may result in serious consequences when critical sensitive equipment are not being adequately protected. Usually, semiconductor plants suffer the most in these type of incidents, as their process equipment etc are very sensitive to variations in the power supply.
Voltage dip mitigation comes in various forms; some are battery-based like the the UPS (which is also a mitigation against a total blackout), while others are batteryless like the SoftSwitching MiniDysc (which caters for variation in the power supply for a few seconds only). The latter is preferred for dip mitigation as batteries require a rigorous maintenance and replacement programme, to ensure the batteries do work when they are called for.
Earlier, our office recorded a ‘shallow’ dip at approximately 1919hrs. Tell-tale signs from other neighbouring monitoring sites show this could be a transmission-level fault. Shall await for the report from the utility tomorrow.
Update 15/9/2014 from the Utility: Customer installation fault at Jurong Island
Noting how this dip due to a ‘customer installation fault’ can be seen in many areas, those ‘in the know’ will know who is the Customer here.
Just awhile ago, if you are one of the facilities’ guys, you probably had to scramble around because of dip alarms, standby generator cutting in or chiller drop-off.
Another 230kV fault was registered; affecting the South side of Singapore the most, suggesting a 230kV fault in the South block.
Here is a voltage waveform from one of our sites in the South.
Voltage waveform is taken at Low Voltage (L-N). It will mirror what is seen at 22kV and above (L-L).
(i.e L1 phase at LV is equivalent to L1L2 at 22kV).
The waveform here shows that there was a single phase fault (L2) at 230kV.
Here it was registered; worst case dip of more than 50%. Other blocks in Singapore (North, West, East) would also have seen this fault, albeit at less severe values.
On a happier note, Selamat Hari Raya Aidilfitri (in advance) to my fellow Muslims. Maaf Zahir Batin.
As a fan of powerXplorer Px-5 and Dranview, I got a little excited when Dranetz made this announcement.
Hope to get my hands on these ‘toys’ soon. 🙂
Go to dranetz.com for more info or visit their video link here.
p/s: If you have not used Dranview before, you don’t know what you have been missing.
In Singapore, a voltage dip is defined as per EN50160; “a sudden reduction of the supply voltage to a value between 90% and 1% of the declared voltage, followed by a voltage recovery after a short period of time; between 10ms and up to a minute”. Typically it lasts less than 200ms. Two things matter in the definition of a dip. See Figure 1.
1) Magnitude: This typically reflects the fault severity and also the proximity of the monitoring point to the fault location.
2) Duration: The timer starts when the voltage falls below the 90% threshold and ends when all voltages are equal to or above the 90% threshold. This is very much dependent on the time taken to isolate the fault and the nature of loads connected.
The Utility company here has a power quality monitoring system in place; from voltages 22kV and up to 400kV. As these are 3 phase 3 wire systems; line to line voltages are used in the definition of a dip. When there is such a dip, the worst dip by magnitude and its duration will be published by the Utility.
There is a significance of such definition being used in the 66kV and 22kV networks, whereby it is a resistively earthed grounded system (thru the neutral ground resistor). Here, a single phase fault will not be a registered as a voltage dip as the other two non-faulted phases will swell; ‘compensating’ the faulted phase. This will result in a drop of voltage (line to line) of usually less than 10% (hence not a dip). The Utility company here described such events as ‘Voltage variation’.
In recent times, voltage dips seem to be quite frequent in our office at 16 Boon Lay Way, Tradehub21. Tradehub21 is one of the many commercial buildings located in this Jurong East region; near to the International Business Park and also shopping malls like IMM, JCube, JEM and Westgate.
In the last 5 months, our office has recorded a total of 3 voltage dips; two originated from the Utility’s 230kV transmission fault and one which occurred last Friday due to a 22kV Customer’s Installation fault (along Jurong Gateway Road). It is fortunate in a way that this area is not a Semiconductor Hub, as just one dip will have costly implications on a semiconductor plant, let alone three.
However, this is not to say that commercial buildings and shopping malls are immune to the effects of voltage dips. Here are the common known effects of voltage dips in a commercial building / shopping mall setting.
1) Chiller System
It is a known fact that chillers plants and its related equipment are sensitive to voltage dips. Relays, control contactors and the electronic/computerized controls may drop off during dips, hence causing the trip. The impact to building occupiers is small; as these effects are likely to be transparent to them; or at most minor discomfort due to a slightly raised ambient temperature.
Possible Mitigation
The sensitivity is sometimes made intentionally (and at times over-conservative) as there is a concern on possibilities of damage to the motor due to the high transient current upon normalization of system voltage (hence it is better to trip). A consultation with the Chillers’ OEM is almost necessary if one decides to protect its control circuits. This is to ensure that it will not result in damaging the motor.
Chiller controls can be secured with on-line UPS or voltage dip mitigation solutions like MiniDySC from SoftSwitching Technologies. Sensitive relays can be identified and replaced. Another option is to enable the Chiller for ‘Auto-Restart’ function if allowable.
Having an internal thorough proper restarting procedure of chillers in the event of tripping due to voltage dips in the network is also a form of ‘mitigation’ as it reduces the downtime of the Chillers. This usually needs close co-operation with the Utility company or install a permanent power quality monitoring system in the building to determine where the origin of the voltage dip (internal fault or fault from the Grid).
2) Lighting Circuits
During a voltage dip, flickering of the lights can be observed. In general, problems only arise when High Intensity Discharge (HID) lamps are used. These types of lightings are known to be sensitive to voltage dips and temporarily extinguish after a voltage dip. Re-ignition time usually takes up to 10 minutes.
Possible Mitigation
For HID lamps applications, one possible mitigation is to re-fit with “hot re-strike” igniter for instant lighting restoration. In the use of HID lamps in “critical areas”, it is best to deploy them alongside normal type of lightings (for eg. Fluorescent lightings), so as to avoid a ‘total blackout’.
In addition, occurrence of voltage dip across three phases is rare; hence the even distribution of lightings across the 3 phases can avoid a ‘total blackout’ during a voltage dip.
3) Escalators
During a voltage dip, the control contactors and PLC of the escalator may drop off. Another cause could be the activation of the phase monitoring relay, which is meant to activate upon loss of mains.
Governed by the Singapore Standard CP15:2004, escalators are designed to be brought to rest in a largely uniform deceleration in the event of loss of mains (thus not a safety concern). However public image may be affected.
Possible Mitigation
The control contactors and PLC can be secured through adding an uninterruptible power supply (UPS) and the phase monitoring relay can be upgraded to a relay equipped with a time delay for up to 0.2 sec. See Figure 6. This is based on CLP Power Hong Kong’s experience in voltage dip mitigation for escalators. It is also documented in Hong Kong’s” Code of Practice on the Design and Construction of Lift & Escalator.”
My personal accounts in conducting voltage dip tests on some of these escalators showed that in general, a voltage dip of more than 40%, regardless of duration that affects the supply to the control circuits will result in the escalator to stop. Mitigation equipment such as the MiniDySC was found to be an effective solution against such events.
A full implementation of such mitigation however will need further tests with the OEM escalator’s specialists to ensure that safety features related to its braking will not be compromised. Additionally, it may also require slight deviations from the existing CP15 as similarly re-defined in Hong Kong’s Code of Practice for Lifts & Escalators.
4) General Circuits
General circuits served by miniature circuit breakers (MCB) have also been reported to have tripped after a voltage dip. Immediately after a voltage dip, some MCBs may tripped due to large in-rush current drawn by motors, switch mode power supply, circuits containing control transformer, etc.
Possible Mitigation
Normally, if the MCBs are properly sized to cater for such ‘inrush’, the circuits should not trip.
SEMI F47
When it comes to voltage dip mitigation, the word “SEMI F47” will most likely to crop up. Without going into details, it is basically a minimum ride-through specification for semiconductor tools and processes as the semiconductor industry are most prone to voltage dips due to the nature of its operations. See Figure7.
However it must be noted that even such specification does not translate to 100% availability during a voltage dip as dips that are required for compliance with this specification occur between one phase and neutral, or between one pair of phase, at a time only. Hence for these 3 voltage dip incidents mentioned, a SEMI F47-rated equipment will also not guarantee you 100% availability.
To put in simply, in the SEMI F47 specification, an economical balance has been chosen such that semiconductor processing equipment which meets the requirements will be immune to most, but not all, real-world voltage dips at semiconductor plants.
Similarly if one is to consider mitigation equipment, a cost-benefit analysis has to be conducted as such mitigation do not come cheap. A historical dip statistics of the area from the Utility Company should always be among the first things to be obtained.
Faults (and hence voltage dips) are inherent in all power systems in the world. The frequencies of such occurrences can be minimized but not eliminated. Minimization requires the cooperation of all Stakeholders; the Regulator, the Utility company, the Licensed Electrical Engineers and Customers alike through regulations, regular preventive maintenance and condition monitoring.
And if need be, voltage dip mitigation equipment can be employed for those with sensitive operations or processes, albeit it will come at a price. But the price of leaving it to chance may even be greater (see Figure 8). Ultimately the user has to decide.