Frequency: Island vs. Mainland

We were looking at some power monitoring data this week for an MRI (Magnetic Resonance Imaging) site located in Puerto Rico. The site power was generally pretty good (in terms of things like voltage regulation, harmonics, sags and swells, etc.) but we were struck by the difference in frequency regulation at this site as compared to a mainland site.

We look at a lot of mainland US power data sets (~300 per year) and frequency is almost never anything but rock stable. Frequency fluctuation is often a good indication of an alternate power source (such as a free-running UPS or back-up generator)

Island Frequency

Frequency at an MRI site in Puerto Rico (one week monitoring, 480 VAC, with good voltage regulation)

Mainland Frequency

Frequency at an MRI site in New Jersey (one week monitoring, 480 VAC, with good voltage regulation)

The frequency was technically “acceptable” at both sites – most equipment is spec’d for 60 Hz +/-1% (59.4-60.6 Hz) or +/- 1 Hz (59 – 61 Hz) and some is spec’d much wider. Switch mode power supplies might work fine from 47 – 63 Hz, for instance.

However, some older equipment does use the line frequency for timing or for regulation, might control output or dose using phase-controlled regulators, or might have some protection or control circuits that trigger off the incoming AC waveform – and these might act up a bit if the frequency varies. So frequency is a good thing to keep an eye on when working with equipment installed on an island or other smaller power grid.

No Current? No Problem!

We recently looked at Fluke 1750 data from an ultrasound site. As is often the case with these 120 VAC, 1 Phase monitoring, it’s voltage only (due to the need to use a break-out cable to monitor current on a manufactured power cord; even if they are made up and sent with the monitoring equipment, they tend to disappear from the power analyzer kits over time.

A lack of current data can be a real handicap when analyzing data, but we go back a few years, to the days before current monitoring was ubiquitous. Here’s a few things we were able to determine, even without current data.

Ultrasound OutageFirst, an outage was captured, identified by the service engineer as “hospital power went out”. We’re not so sure; the rapid loss of voltage, with no significant decay, points to a local event (not facility wide), and the neutral-ground voltage swell and phase voltage collapse points to a high current event. Our best guess – equipment fault caused an overload and a subsequent breaker trip.

Ultrasound TransientsNext some voltage transients, captured a few times during monitoring, on both phase-neutral and neutral-ground. Without current data, it’s hard to make much of a guess regarding the source of these. But zooming out a bit provides additional evidence.

Ultrasound Switch-onRMS voltage logs show some small drops in voltage immediately after the transient – typical of a load switch-on (neutral-ground data, not shown, also supports this). We’re sure the transients are associated with load switch-on; probably switch bounce of a circuit breaker, relay, or contactor.

Power analysis without current data. Kicking it old school…..