Editorial note: This blog post is from August 28, 2016. Fenton has arrived home safely!

by Fenton Rees

As I (Fenton) am wrapping up my 8+ weeks here (I leave here Tuesday AM), one interesting thing to contemplate is that if I have been even somewhat successful, the results here will be invisible. If I have been successful, people won’t get electric shocks anymore, there will be no further electrical fires, the generators will no longer get overloaded, and medical equipment will no longer get damaged from voltage surges. It will, in a word, be completely uneventful--but obviously in a good way. That is in contrast to Pat, for when she is successful, someone with a previously fatal ailment walks out with the rest of their life ahead of them.

I suppose my results being largely invisible is similar as to how things work in the spiritual realm, where we are aware of only a tiny fraction of the results of anything we do for the Lord, even those that are very visible such as preaching or teaching.

Today (Saturday), Caleb and I got to try out something we have been working on for quite a few days. We were looking for a way to stop all the water heaters in the mission houses from overloading the generator and the new voltage regulator when the power comes back on in the morning. The water heaters would cause stress on the generator and voltage regulator because when the power system was reactivated, the water heaters all turned-on and would increase demand for electricity as they had cooled off overnight. I borrowed an idea from the electricity grid in New Zealand, where fifty years ago both water heating and most house heating was electric (now there is natural gas). To reduce the peak electricity demand in the middle of winter, the New Zealand grid sends out a signal over their power lines which turn off all the water heaters in a given geographic area for an hour or so. The next hour is then a different area and so on.

We could not be so sophisticated. We had to bury a signal cable to each house that operated a relay that controlled the water heater. That required about four hundred meters (one quarter of a mile) of trench to be dug, about one-foot-deep and four to six inches wide. Caleb had about twenty guys out there digging away. In less than a day it was all done. One thing these people know how to do is dig. Apparently, one of the first verbs you learn when studying Kirundi (the local language) is hoeing or digging, and when you are here it is obvious why--both men and women know how to--as that is how most feed themselves.

I had initially thought that using some kind of wireless connection to each water heater would have been better, but labor here is so cheap that digging a ditch was by far and away the best way to go. We spent about one hundred dollars on the four hundred meter of wire for the control signal, about two hundred dollars for the relays, but one dollar per day for unskilled labor. The four hundred meters of trench cost us less than twenty dollars.

Now (Sunday), after twenty-four hours of operation with the new relay system, we appear to have success.

And for those of you who don’t care for the technical stuff, Pat has some words for you:

I was pondering today emotions that I see here in Burundi. The Burundians don't cry and they don't raise their voice in anger. They often laugh when we as Americans would be weeping or raging. They don't look you in the eyes, out of great respect. Yet, when you go to talk to one patient about an issue, you will look up and everyone in the room, the hall and from outside will have come to gather around the bed to listen, even if they don't know the patient you are talking with. They often say "you must do everything" when you are at the bedside of someone about to die, but will question if they need to really drain some obvious large pocket of pus on their small child, as they may not have the ten dollars for the operating room fees. Emotions and cultural concepts I don't understand.

However, there are things I do see that are universal and beautiful here:

Saturday afternoon I was in my office reading and a line of people were waiting to see the medical doctor next door. A baby began a sad and sick type cry. Soon I heard the mom crooning and patting the baby, and I peeked around the corner to see her in the typical position of rocking back and forth with her babe in her arms and attempting to nurse the child to quiet and console the wee one.

Tistou is a medical student from Hope Africa University in Bujumbura. Although Tistou grew up in Mozambique, he is Burundian who is fluent in English and Portuguese. He came here to go to school and his father had died before he arrived. His mother died two years ago, as did another brother. He is the oldest and therefore responsible for his younger brother in high school. He had no money to continue school but did get some scholarships to continue. He does extremely well clinically, but he tells me he has had difficulty with classes, mostly because he can't see the lecture material as he needs glasses and couldn't afford the 90,000 BFU ($45) to buy them. His personal family loss gives him amazing empathy with patients, and I have seen him encouraging a little nine year old girl to try to eat when she won't eat for her family, and I have seen him sitting and comforting an eighty-five year old man who was trying to decide if he wanted emergency surgery or not: Compassion given to strangers in Jesus' name.

The surgical ward is full of children with chronic wounds from infected bones and muscles that have harbored pus. Yet, even these kids form a kinship around the crutches, walkers, and occasional wheelchair that they can find to move them from their beds to the small, open courtyard in the sun. There they will be in a line, obviously consoling each other as they each know the pain of their disease and the misery of dressing changes endured once to twice a day.

The love and compassion flow here, even if it is expressed differently than in the America. I believe that this is a reflection of the mission of this hospital "To work with the One who heals and gives life". It flows from the medical staff, the nursing staff, the students and even those that are serving by cleaning and carrying. Love was the message of Jesus. I pray that it is the main message I leave here.

Fenton again:

I head home on Tuesday, and Pat will be here for another three weeks. So pray for my travels (I think the total “flight time” is thirty-two hours including stop-overs), for Pat as she stays here, and for the hospital and full-time missionaries who will still be here even when we are both back home. This would not be an easy place to live year after year.

For those of a technical “bent”, below is a summary of the electrical work I have been involved with for the last 8 plus weeks. I’ll spare you the other fifty plus pages of details.

2 Months of Engineering at Kibuye Hope Hospital, July | August 2016

Aug 28, 2016

Fenton Rees at Kibuye

Contents:

1. Introduction.

2. Summary.

3. A Survey of Existing Wiring Practices.

4. A Review of the Grounding Practices.

5. A Partial One-Line Diagram and the Circuit Breaker Issue.

6. Campus Electrical Load Measurements via Data Logger.

7. Installation of the Eaton Power Conditioners.

8. The Fluorescent Light Problem at the Hospital.

9. The Water Heater Problem in the Professional Housing Area.

10. The Professional Housing Area;- Future Plans and Issues.

11. Use of the Professional Housing Area’s Generator to back-up the Hospital.

12. The Water System’s Problems.

13. The Hospital;- Future Plans and Issues.

Appendices:

3-1: Sanity-check by eMi of Wiring Practices Comments in Ch1.

3-2: eMi Wire Chart.

7-1: “Nasty-Gram” to Eaton re poor Design of their Power Conditioner.

7-2: Eaton’s reply to questions about PF caps, 3rd harmonics, fuses and over-voltages.

7-3: Mathcad Spreadsheet to Optimize the PF Cap Rating.

7-4: Mathcad Spreadsheet for L-L Voltage Variation vs. Phase Angle Changes.

7-4: Constant Voltage Transformers 101.

8-1: Mathcad Spreadsheet for 3rd Harmonics in the Neutral of a 3-Phase System & Experimental Measurements.

10-1: Current Limiting by the Fast Fuses at the Output of the Eaton Power Conditioner.

13-1: Solar Power at Kibuye?

13-2: Proposed new Electrical “Back-bone” at the Hospital.

13-3: Mathcad Spreadsheet for an Optimum Air-cored Inductor.

13-4: The utilities

Ch 1: Introduction

In 2011, my wife (Dr Pat) and I became acquainted with Dr Frank Ogden (thru a mutual 3rd party), and were persuaded to visit Kibuye Hospital and to help out as needed - she on the medical / surgical side and me with maintenance and engineering. Suffice it to say that things were a tad on the sad side in those days - no running water in the hospital, no back-up generator, no ex-pat staff of any kind, and no one who was really functioning as the head of maintenance & engineering. Dr Frank had retired from Kibuye some years before due to medical issues and age and was hoping that all his efforts over 20+ years did not just run down to zero. Pat and I then visited Kibuye again in 2012 and she by herself in 2013. For me, that 2nd visit was a little disappointing, highlighting how poor the maintenance was (due to the lack of anyone who had a clue), and confirming in my mind that at most all that short-term visitors could do was to fix a few broken pieces of equipment. Planning for any long term improvements was at that time completely out of the question.

In 2015, I spent a total of 2 months at Bongolo Hospital in Gabon - initially in February with a design eMi team and then later (with Pat) in the year (Sept / Oct) to help install a CVT (constant voltage transformer) at the eye clinic that they had purchased on the recommendation of the eMi team. Prior to the eMi team’s visit, they had experienced the “usual” slew of electrical fires, electrical shocks and damaged medical equipment from voltage surges etc.

So, then to 2016, when Pat became acquainted with a need for surgical cover at Kibuye while Dr Jason Fader was on furlough for 3 months. I think initially the awareness of the need came via Samaritans Purse and then later directly from Jason. Pat wanted me to go, but I needed to be convinced that there really would be something useful I could do - based upon my prior experiences in 2011 & 2012 I was a tad dubious. Two things convinced me that this would be worthwhile use of my time - there was now a full-time engineering / construction / maintenance person (Caleb Fader), and they were about to install two new power conditioners to help stabilize the electrical system. Both of my consulting clients had gaps

in the projects I was working on, so the Lord seemed to open the way. So here we are, on-site at Kibuye from June 29 thru the end of August for me and Sept 21 for Pat.

This document captures the key points of most of my work and “thoughts for the future” from those 9 weeks - hopefully at least some is/will be beneficial. The simple fixing of assorted equipment is not covered. You are welcome to e-mail any criticism, suggestions or questions related to this document to FentonRees@ieee.org.

We appreciate the great treatment we received from the “Serge” mission team and hospital staff. We pray that as the physical infrastructure and missionary medical staff is built up, that much more than that is happening behind the scenes.

We pray that Kibuye Hope Hospital in Kibuye and the associated Hope Africa University in Buja will be agents of change not just for medical care, but for transformation of Burundian society as a whole. May God raise up Burundian versions of Billy Graham and William Wilberforce who will affect improvements to more than just health care. After all, Burundi didn’t get to be one of the 10 most wretched countries in the world by JUST having poor health care. There are all sorts of other problems where the light of the Gospel needs to penetrate as well, including the ineffective and corrupt government, bribery, famine, poverty, and violence. But hopefully the Body of Christ in Burundi will be the solution to all these problems, even if more slowly than we would want.

Ch 2: Summary

One can gather and generate a lot of info in 9 weeks. Below is the “Reader’s Digest” version of the key points. In both this summary and the main text, an attempt has been made to make the info understandable both to EEs as well as other technically minded people - so there is a little more explanation than if it were for EEs only.

As a result of this work, a substantial “to do” list has been generated. This list needs to be prioritized so that:

The risk of electrocution is taken care of first (ie. from lack of suitable grounds). The fact that no one has ever been electrocuted here at KHH does not necessarily predict the future if nothing is done - more medical professionals and more medical equipment dramatically increase the exposure to possible risk.
The risk of fire is then next, (from poor wiring practices) - where the assumption is that the risk of actually dying in a fire is not that great as most of the buildings are built of masonry, are not that large and typically have many exits.
Then all the other stuff can be done - such as load management that will stop generators or power conditioners from being overloaded etc.

Ch 3: A Survey of Existing Wiring Practices

Problem: A “boat load” of serious issues; capable of exposing people to the risk of electrocution from lack of grounds, and of significant fire risk due to inadequately protected cables and wiring.

Solution: An extensive review of the existing wiring practices was conducted. The recommendations were then sanity-checked by eMi-USA. Some of the most important improvements that are now in process include:

Using proper aluminum-compatible lugs when mixing copper and aluminum cables at the same terminal. (Some will arrive with the ITEC team in Sept).
Adding ground wires to outlets where none existed - completed at the old school and cottage, but NOT YET DONE at the Ogden house, which is therefore a potential death trap.
Right-sizing circuit breakers so they are rated at no more than the current rating of the downstream cable - still far from complete.
Checking that all wire and cable used is actually solid copper and not copper-coated steel; as is present in the Quad-plex, the McLaughlin house, and at last partially in Allyssa’s house and should be replaced ASAP.!! In the interim, all Quad-plex / McLaughlin Alyssa circuit breakers should be replaced with 6A ones, except for water heaters (10A). Not yet done!! NOTE that some stainless steels are non-magnetic, so non-magnetic does NOT automatically mean solid copper.
Ensuring that the ground pins of outlets are actually connected to ground - especially in areas like kitchen and laundry where appliances with metal exteriors are present. Given the poor quality of the Chinese made outlets, this should be a CONTINUING review that is carried out every 6 months.
Continue training and monitoring the current “electrician”, who at least some of the time appears to be guessing as to what to do.

Ch. 4: Grounding Issues

Problem: A ground resistance tester was rented for a nominal fee from eMi. This showed that the resistances of the existing ground rods (to ground) was between 250 and 1500Ω - compared to the acceptable per-building value of <25Ω. Hence there is both a human safety and equipment safety issue.

Solution: The 1.2m long copper-coated ground rods available in Buja appear to be OK to use - it is the surface of the rod that really matters, not the interior. 60 such rods were purchased (~$10 ea). Before the ground tester had to be returned to the US, an initial trial of 6 houses connected as a Multiple Earthed Neutral system was conducted by:

Adding 4 ground rods to each house.
Connecting the new ground rods to the house ground and connecting that to the Neutral.

Measurement showed that the combined ground resistance of that whole assembly was then 13Ω - a dramatic improvement. This has all the rods working together via the neutral conductor, “a cord of many strands is not easily broken”.

Continued Work: 4 such rods should then be added at each remaining building that contains either expensive or sensitive electrical equipment: so all doc houses, OR, eye clinic, and anywhere else that seems reasonable. The resistance of the complete assembly will then hopefully be a very respectable 5Ω or less. This is only partially complete.

Ch. 5: A Partial One-Line Diagram and the Circuit Breaker Issue

Problem: A partial “one-line diagram” of the Professional Housing area highlighted that the over-rated circuit breaker issue extends all the way through the system - from the hospital’s generator shed to the individual houses / buildings.

Solution: Demonstrate that even when circuit breakers are working per specification, they trip off on mild overloads (ie. 1.5x to 2x rated current) much more slowly than you would think. ie. A 6A breaker loaded to 9.4A (1.57x rated) took 7 mins to open! And as the heating (and temperature rise) in a cable is proportional to the SQUARE of the current, overloading a cable can quickly result in a problem, which is hugely aggravated by over-rated circuit breakers.

Continued Work: Suitably rated circuit breakers have been acquired but their installation is still not complete.

Ch. 6: Campus Electrical Load Measurements via Data Logger

Problem: Almost no information was available as to the currents that were drawn by the various parts of the campus or the campus as a whole - just the occasional measurement via hand-held ammeter.

Solution: Used a relatively inexpensive (~$200) 2-channel data logger to measure the voltage and current of one phase of the total campus for a week. The other two phases were then done in subsequent weeks. So this included the hospital, professional housing, water pump and a few other small buildings, but excluded “new Busoma”.

Results: It was found that:

The grid is often off from late evening to around dawn, (We more or less knew that). Although as we progressed thru July the grid outages did become more frequent, to be expected as dry season progresses.
Max and min voltages of 423 and 366v were seen - within a +/- 10% tolerance, so not catastrophic. But as the sampling was only once per minute, much higher shorter duration surges were probably not captured.
Maximum currents in the region of 75 – 85 amps was seen on all 3 phases;- once when the grid came back on in the morning and then again after dinner time around 8pm.
A quick check with a clamp-on ammeter showed that the evening’s current peak did not come from the hospital but from the prof. housing area;- most likely the water heaters kicking on after dishes or kids are washed.

Conclusions: Even if those 75 – 85A current peaks were split evenly between hospital and housing, that means that each will see 37-42A per phase;- worryingly close to each back-up generator’s rating of ~40A or each Eaton power conditioner’s rating of 43A. See Ch8 and 9 for ways to reduce the loads at both the hospital (by using LED lights and/or power factor correcting capacitors) and professional housing (by controlling when the water heaters come on).

Ch. 7: Installation of the Eaton Power Conditioners

Problem: Two Eaton power conditioners (= voltage regulators) were on-site waiting to be installed, one for the hospital and one for the professional housing area. These are specified to limit the voltage variations to ±3%, which should help protect delicate medical equipment and household appliances. The challenge was installing them so that they would be affected as little as possible by the less than ideal conditions;- endemic rats (both units were infested by rats during on-site storage) and lots of dirt and dust.

Solution: For the unit for the professional housing area, a new shed behind the generator house was built, hopefully rat proof and somewhat away from the dust of the nearby road. For the hospital, the unit was installed in Rm 10 nurse’s station, so access to the main hospital panel out behind was nearby. This office is clean enough but not yet remotely rat-proof, a major concern. Additionally, as these machines are relatively complicated and therefore could potentially fail, they were installed with a transfer switch which would allow the unit to be bypassed in case of failure, and unregulated grid power used.

Sanity Check Operation: As these power conditioners were new to everyone involved, their operation was investigated to make sure there would be no unpleasant surprises. The prof housing’s unit operated flawlessly even when subjected to Caleb useing the arc welder in the Container-Plex;- checked both by data loggers an scope. But a week’s worth of data from the hospital’s unit showed that the current maximum load (38A) is already very close to the unit’s rating of 43A per phase.

L-L vs. L-N Voltages: This Eaton voltage regulator electronically changes the taps on the three individual phase windings of the transformer’s secondary. It was found that the measured voltage stability of the L-N voltages was substantially better than the L-L voltages. At both the hospital and housing area almost all the loads are single phase, (ie. L-N), so the higher than expected L-L voltages that have been measured are not too much of a worry.

Further Analysis Suggests a Means to Reducing the Load on the Hospital’s unit: Oscilloscope measurement confirmed the unit’s read-out that the power factor is sometimes quite poor, at 0.57, (Bad is 0. Good is 1.0), at least somewhat due to the harmonics generated by the approximately 200 fluorescent lights. But “power factor correcting” capacitors could be added and would likely knock about 9A / phase off the max load seen by the generator and/or power conditioner. It was determined that a 3-phase PF capacitor rated at 15kVA was the best fit for the measured loads and power factors; and such a unit has been ordered and will arrive with the ITEC team in September.

Needed Spare Parts: As part of their protection system, these Eaton power conditioners have “fast” fuses at their output. These fast fuses will clear a short-circuit much faster than a circuit breaker, and so are likely to blow FIRST when there is a significant short circuit in a house. Consequently 10 of these fuses have been ordered and will arrive with the ITRC team in Sept.

Remove Exiting 1-ph Voltage Regulators: Now that we have regulated power, REMOVE all other regulator boxes;- they are relatively unreliable and can cause problems;- like the one at X-ray with its output stuck at 267v !

Thoughts regarding Further Voltage Regulators for New Hospital Buildings: Constant voltage transformers have advantages over the above Eaton type electronic tap-changing voltage regulator, and should be considered for the likes of the new eye clinic or new OR;- where ultimate protection of expensive medical equipment is a priority.

Ch. 8: The Fluorescent Light Problem at the Hospital

Problem: The data logging in Ch6 already identified that the maximum current drawn by the hospital is quite close to the rating of both the generator and the Eaton power conditioner;- and at least half is that from the ~200 40w fluorescent lights. When the new B19 ward building opens in Q2 ’17, its additional 100 lamps (plus other stuff) will mean the combined total almost certainly exceeds the generator and/or power conditioner’s continuous ratings.

Possible Solution: Change the 40w fluorescents to like-brightness 16w LED lamps;- reducing the current drawn per lamp from 0.28 to 0.11A. This would knock 17A per phase off the max current drawn, (that’s existing hospital plus B19), and allow the existing generator and power conditioner to be continued to be used.

But this Solution has a Problem: LED lamps operate from DC power and the rectification process to make DC from AC generates “harmonics” that can sometimes cause bad side effects. These cheap Chinese LED lights (and ceiling fixtures) are probably worse in this regard than better made ones sold in the US. Measurement of their current waveform (via Laptop scope) showed that the 3rd harmonic (ie. at 150Hz) so generated was significant, especially on the neutral where it gets amplified. Eaton confirms that their unit is NOT rated for significant amounts of harmonics.

Hence the easiest solution (LED or fluorescent) is to leave the existing hospital as is but for all new buildings, (incl. B19), to have say at least half of their lamps fed directly from the grid, and not via the Eaton power conditioner;- reasonable as the lamps are not that expensive if taken out by a voltage surge.

Ch. 9: The Water Heater Problem in the Professional Housing Area

Problem: As described in Ch6, the peak load in the housing area is approaching that of either the generator and/or power conditioner, and is mostly due to the water heaters all coming on at the same time. The addition of TWO more water heaters at the Baskin house when it comes on-line in a month will only make things worse.

Solution: Add control relays at each water heater (<$15) and control via an underground control wire that allows only half of the water heaters at a time to come on. As the hot water tank serves as a huge “thermal battery”, continuous power is not required;- measurement showed the water heaters can heat from cold in 1½ hours. Hence cycling the hot water heaters on for 30 mins and off for 30 mins should not be especially inconvenient.

Results: After installation of the above system, a few days’ worth of data logger info showed that the maximum current drawn from the prof housing area’s Eaton voltage regulator was now only 25 ? amps, confirmed by the readings on the Eaton’s front panel. While this is with some families out of town, it suggests that we should still be below 40A even with all planned additional houses completed and occupied.

Before the above timing circuit was operational, an experiment at the McLaughlin’s house showed that manually turning the hot-water heater on for just 4 hours per day versus leaving it on and allowing the thermostat do its job saves so little money (~$3 / month) it is a waste of time. Hence the dedicated hot-water switch on the wall can be eliminated in future builds, and the circuit breaker used if the water heater needs to be turned off when the house is vacant.

Ch. 10: The Professional Housing Area | Future Plans and Issues

Question: Will any kind of solar energy / battery system ever make sense for the professional housing area ?

Answer: If an absolutely worst-case scenario is imagined, (housing area fully built out with about 20 units, almost no grid power, diesel at $3 / liter, and run the generator for 4 hours a day at near max load), then the fuel cost per housing unit is about $180 per month. No more than an average American electricity bill, and probably tolerable.

Easiest Means of Reduction: But as at least half of that is for hot-water heating, and under those circumstances it would make sense to add solar hot water to each house. The payback time would be about 2 years. But of course if both electricity and fuel stay “cheap”, (~US10C per kw-hr and ~$1/liter), then the payback time will be 3x as long. Possibly solar hot water heaters should be added to all new houses.

Safety of Circuit Breakers: The existing utility distribution transformer (estimated at 150kVA) and the long cable length from it to the professional housing area means that the maximum possible short circuit current is substantially below that which would damage the smaller (< 100A) circuit breakers.

Ch. 11: Use of the Professional Housing Area’s Generator to Back-up the Hospital

Problem: If the hospital’s generator fails in some way, then it would be nice to somehow be able to supply thehospital from the Housing area’s generator.

Solution: Preferred solution;- add a buried “status” cable between the two generator shed’s so that they can be properly interlocked to prevent bad connections being made (eg. grid connected to generator). 2nd best solution (not yet implemented) ; add four transfer switches and have procedures in place to “lock” the two most critical switches to prevent unauthorized operation.

Ch. 12: The Water System’s Problems

Problems: 1 ) The pump controller is near the pump, so when it trips off from under-voltage no indication is available that this has happened. 2 ) There is no level sensing at the water tank, so the length of time required to run the pump to fill the tank is just guessed at.

Solutions: 1 ) Move the pump controller to the hospital’s generator shed (done) and add an alarm lamp (done). 2 ) Not yet done;- add level sensing at the water tank and connect via buried cable to the pump controller in the generator shed. Or possibly use a pressure sensor in the feed line FROM the water tank to estimate the water level and eliminate the buried cable.

Possibly the use of the relays on the housing area’s water heaters will sufficiently reduce the peak load on the Eaton power conditioner and generator to allow the water pump to be powered from it, and hence eliminate its under-voltage issues.

Ch. 13: The Hospital: Future Plans and Associated Issues

Existing Status: As of Aug’16, the hospital’s generator and power conditioner are just making it during periods of peak load. As grid power is relatively cheap and mostly available during the main hours of operation (7am – 10pm), and as diesel for the generator is relatively cheap, if this status was to remain, there would be no good reason to contemplate solar or battery power, (Unless perhaps donated by a sympathetic tree hugger). The size of the utility transformer and distance from the hospital provide adequate short circuit protection for the smaller lesser-rated circuit breakers.
Future Changes: Major changes to the campus will occur in the next 5 years;- many more buildings, more equipment and a probable change in the location of the main electrical distribution center / generator. Outside changes that could occur include dramatic price increases in both grid electricity and diesel fuel, and the addition of a new dam 20km away that perhaps will make the grid more reliable. Even with all these campus changes, if fuel and electricity stay cheap, there appears to be no overwhelming reason to contemplate solar or battery back-up;- unless of course it comes for “free”.
eMi Master Plan (June 2015) with 12 distribution transformers: This plan, at least partly driven by the need to have sufficient impedance in order to protect smaller down-stream circuit breakers, appears to this author to be the wrong solution to the problem, for two reasons:
The multiple smaller transformers mean there are more connections to the 30kV grid, versus fewer connections with fewer larger transformers;- so more chances for problems that require the somewhat unresponsive utility to fix.
It creates many more individual “islands”, so that backing up via generator and/or with solar & batteries would require a back-up system per “island”, versus one larger generator or solar system.
Alternative solution: A single larger transformer (eg. 500kVA) could be used for the entire hospital complex at its maximum build-out, allowing a single larger generator and single solar/battery system to be used. Each smaller feed (say 100A) to the individual buildings could have some added line inductance (= 1.5% impedance) to protect the smaller down-stream circuit breakers from excessive short circuit current.

Acknowledgments:

While I hae done most of the measurements and calculations, and written this report, a whole host of others have contributed mightily, including:

Caleb and Tony and the maintenance/construction team from buying supplies in Buja, to having the house built for the professional housing area’s power conditioner, to digging holes and ditches for ground rods etc..
Friend Dave Oleson and brother in Australia (Ian) for sanity checking this summary and helping me both simplify and bring focus to unfinished business.
Andy Engelbretson and Hannah Paterson at eM-USA for both sanity checking some of this work (especially the wiring practice Issue in Ch. 1 and thoughts on use of the housing’s generator by the hospital) and the letting me have the ground resistance tester for a nominal fee.
Ron and Dean at Anderson Electric Controls (Kent, WA) for helping get product info on assorted components and buying the spare fuses and PF caps for the Eaton power conditioner and the relays and timer for the remote water heater controls.
John Joiner for reviewing the installation and testing of the Eaton power conditioner and making some other helpful suggestions.
Paul Taylor in Bend for helping with info on the Eaton power conditioner.

All for now!