*System containing UV lights for water treatment
Summary:
The electrical grid powering Our system at Davis Dam is characterized by high inductance and low impedance, with frequent voltage dips - probably due to electrical power load changes.
We observed many voltage dip events in the power grid. We identified four different types of dips based on their duration – ranging between 2.5 msec to 2.5 sec - and the percentage of voltage drops.
We did not identify the source of the dips during our visit and therefore cannot recommend how to prevent them from occurring.
One of the four types of dips happened every 40 minutes: we did not observe any influence on our equipment.
Two of the four types of dips caused the input capacitors of the ballasts to resonate with the grid inductance, leading to extensive current ringing with a resonance frequency of about 1550Hz; we did not observe that those types of dips were causing damage to our equipment. .
The fourth type of dip (50% voltage drop with a duration of 2.5msec) created, upon interaction with the ballast, high phase currents. These currents rectified inside the ballast and rapidly change direction, causing voltage surges that led to ballast failure - Mains Fail.
2
Based on our measurements at the site and lab tests we conducted, we expect that in the event of surges that follow the line grid dip, the newly installed Power Supply (with filtering), with high confidence level, will limit the surges so that no damage will occur.
In case this solution fails, we will probably need to isolate Our system from the voltage dips by means of a separate clean line or by using UPS isolation.
In the figure we can see Grid current and Voltage
In this figure we see the Commutator current
We know that high inductance will cause smaller damping so the circuit will ring more before reaching stabilization point. We also so know that higher capacitance and impedance improve the damping time.
At figure below you can see the damping at the system at Davis Dam while working with 2 ballasts (L=200uH, C=18uF) during voltage dip
The data logger which was installed at the site at the 4th of November shows that there are many voltage dip events at the power grid. We observed so far four different types of dips:
We will now show detailed graphs of events 1-3 which so far proved to be not harmful to the ballasts (event number 4 which was harmful will be detailed at the next paragraph):
1. At figure below you can see the RMS values (voltages and currents) of type #1 event and at next figure you can see the real time values (voltages and currents).
You can see that these minor dips events don’t cause any fluctuation at the input current. We now estimate that this kind of event cannot cause damage to the ballasts.
At figure below you can see the RMS values (voltages and currents) of type #2 event and at figure 6b you can see the real time values (voltages and currents).
At this kind of voltage dips there is high current fluctuation which put some stress on the ballasts.
The event at the 4th of November which captured at the PQube data logger caused failure to all 6 ballasts and also to the PSU protection device which was installed at the Systems.
A new surge protection device was designed for implementation inside each one of our ballast modules to prevent surges to be missed by the PSU and break the chain reaction in case of an individual failure. This protection device increases the system stability to line fluctuation.
Follow up:
It is yet to be seen whether the PSU is capable to absorb energies occurring in the system. However due to the number of components, the PSU also increases the impedance and therefore increases the damping factor in the event of a dip. The PQUBE data logger will record coming events and document how the system is dealing with it.
Image of the PQube3 Electrical network logger and Analyzer
