Wild weather in Tuvalu

At about 10:30pm Monday night it hit. We’d been waiting for it since about 7; the promised high wind/cyclone that has delayed our ship yet again and has left us all packed up with nowhere to go.

We’ve been working 6 ½ day weeks for the past 4 weeks with a goal in mind to be finished here and ready to move on to Nanumaga in line with delivery of our equipment and ferry schedules. After all the delays in getting our equipment unloaded here, and the realisation that logistics will be one of our biggest challenges out here, it seemed sensible to work harder and faster, giving ourselves the biggest possible window filled with the most potential transport options to move on.

On Thursday morning we had news that the Komaiwai was unbelievably running ahead of schedule and might be ready to sail past Vaitupu for a pick up as early as Saturday morning. All our hard efforts seemed worth it. We could be finished. We could be ready and packed in time. We worked until 11pm Thursday night and were back on site by 7am the next morning.

And the site looks incredible. The new Vaitupu powerhouse is an impressive piece of engineering; the result of months of planning, agonisingly detailed calculations, measurements and drawings and a shipload of the finest German technology that money can buy.

The finished inverter corridor in the powerhouse

Amidst the frenzy of finishing last week, we opened its doors to the local community. On Thursday the primary school students and their parents came to visit. Still dressed in our sweaty, grubby work clothes, we were presented with beautifully made floral wreaths and a woven basket of coconuts that we devoured within minutes of their all heading home for lunch. 

On Friday morning the high school students arrived; ferried in by their Fijian teachers and primed with good questions. Thankfully no-one asked if the generator amp setting on the master Sunny Island was drawing the correct current from the generator… we were still tinkering.

Fifty visiting 16 and 17-year olds crammed into the battery room

That night we celebrated everything we’ve achieved here so far with a spectacular feast. Hosted by our wonderful Fijian housemates, we ate crab and reef fish and breadfruit cooked in coconut milk and chilli and fresh made roti and a local fern, which has become our primary source of greens. The boys were dressed in brilliantly loud and colourful shirts and I finally had an excuse to bring out the Tuvaluan outfit which I had had made in Funafuti. As the night warmed up and our gin ration stretched, we sang songs of farewell and safe passage. We have been promised that if PowerSmart comes to do a project in Fiji, we will be feasting every night.

Celebrating with our Fijian housemates

And so like sad prom dates, we are waiting at the curb for the limo. And a cyclone is coming.

It rained horizonally for most of the day on Tuesday. The sparser of the coconut trees by the wharf, trying to stand solo against the wind, look weary, and someone has started moving boats off the harbour ramp. Apparently one was swept up on the water and came crashing down again. It has holes.

By 8:30am Wednesday morning, it’s high tide. There is a crowd gathering in front of the meetinghouse above the wharf. Some of the locals report that they’ve not seen weather like this since the storm of 1991. Waves have been pushing higher all morning sweeping rubbish up onto the streets. Two coconut trees have given way and come crashing down. With each set that comes in, the marker pole at the end of the wharf disappears in a spray of white and another wall of blue appears just behind it. There is no conceivable way that a little barge with a 40hp engine would make it through that harbour entrance unscathed. On a day like today, no ships will even be allowed to fix to Funafuti wharf; they are bobbing like corks in the middle of the lagoon, waiting just like us.

Storm watchers

Vaitupu harbour under siege

Assessing the storm damage

Easy to reach these coconuts!

Weather report released Monday night…

Tuesday and Wednesday: North to north-easterly winds 15 to 25 knots gusting to 30 knots over open waters. Seas rough with westerly swells 3 to 4 metres. Thursday? Westerly swells 4 to 5 metres.

This afternoon we wanted to see for ourselves what damage this weather is causing. As we head to the western tip of the island we see pig pens washed away and outdoor kitchens being reinforced in anticipation of high tide which is due tonight. At the tip of the island, we traipse through the mud and peak out through the Pandanas palms. Something doesn’t look right. I realise that the horizon isn’t flat. It looks like a rolling mountain range; all dark green with snow-capped peaks. There is a point on the horizon where the very western-most corner of the reef is reaching out into the ocean and a great pyramid of water is pushed into the air. The wave peaks and breaks from this point, rolling toward us. I’ve never seen waves like this. My sandy, coconut tree covered, 2m high atoll, perched in the middle of this enormous ocean, feels very precarious right now.

Team outing to check out the waves on the western tip of the island

And to top it all off, we are on diesel rations again. We are down to our last 235L of fuel and the generator will only come on for 3 hours tonight and then another 3 hours tomorrow morning. With it our communications with the rest of the world as the satellite uplink relies on the generator and reasonably clear skies. I will try to get this post out tonight but it may be a few days before you hear from us again. It’s time to hunker down. To drink too many cups of tea and spend half the day scurrying to the bathroom. Or perhaps to bring out the scotch whiskey and work on our fantastical stories of the Tuvalu storm of 2015. 

Building a mountain of empty diesel drums

Solar 101

Shane has been running night classes for local workers and the technicians who will be responsible for managing the system when we leave. It’s been a chance for even his team to further grow their understanding of what we are building here; how the component elements fit together like building blocks to efficiently deliver the energy capacity that is needed here. We’ve been asked by a few people to put it in writing so this knowledge can be shared. So here goes the not-too-technical guide to off-grid solar. Or the long answer to the question “so what exactly are you doing out there?”

Electricity flows like water through a system that allows you to store reserves and pump it in different directions to manage availability and meet demand. This is the starting concept for class #1.

 

Reusing panel boxes as educational supplies

Just as rainwater is captured, pumped into tanks when it’s raining and then pumped in the other direction when it is dry, so energy can be transformed and moved to where it is needed.

In our system, solar panels “catch” the sun’s energy like buckets in the rain. This energy travels along underground cables to the solar inverters that convert the energy from direct current (DC) to alternating current (AC) that we use in our homes. All this energy then goes to the multi-cluster box (MC Box) which acts like a big controlling valve at the heart of the system.  If the MC box is the heart the system the battery inverters are the brain. And one battery inverter rules them all. The master battery inverter decides where this energy goes. Direct to the village? Or if the sun is blaring and more is being produced than used, excess energy can be sent to the battery inverters that transform it back into DC, charging the batteries, which act as tanks.

At night when the sun goes down and everyone turns on their lights and start watching Masterchef Tuvalu, the master battery inverter changes the direction of flow and starts drawing that excess stored energy from the batteries and using it to power the village.

And if the Tuvaluans are unlucky enough to have a few days of torrential rain with no sunlight, the master battery inverter is clever enough to send a message to start the generator, which acts like a pump drawing another source of energy into the system.

How your Solar electricity plant works

So how will having this system change energy use on Vaitupu?

Vaitupu is currently powered by a 110kW generator, which operates 18 hours a day (from 6am to 12 midnight). Over this 18-hour period, this hungry beast consumes 250L of diesel. Over a year that’s enough for me to drive my car 1.2 million kilometres (or 60 people to drive their car 20,000kms)! And that doesn’t include the energy used to transport the diesel to this tiny remote island.

And while the village load rarely exceeds 50kW at any time, a generator this big is needed to manage peak demand in the morning when everyone’s freezers go into over-drive and to provide some redundancy. So there is a lot of diesel being burned and not all of the energy that is being produced, can be used.

The current energy load profile for Vaitupu is hinted at in paragraph above. At 6am every morning when the generator comes on, all the fridges and freezers that have slowly been thawing over the past 6 hours, turn back on. And they have to work very hard for the next few hours to cool everything back down again. Things start to settle around 11am and then there is another peak in the evening when everyone gets home, starts cooking dinner and turns the lights on. The current load profile from an average day is plotted below.

Average daily Vaitupu village energy use

Once the solar system is turned on, it is expected that the load profile will level out. There will still be a bump in the evenings but no more morning peak and less work for all those exhausted refrigeration motors; fridges will now come on for a few minutes every hour; food will stay frozen; fans will keep running overnight keeping mosquito’s at bay and visiting Kiwi’s will be able to sleep.

Back at the powerhouse, the battery state of charge will slowly go down overnight. And then when the sun comes out in the morning, the solar array will start pumping energy into the system fuelling the village and simultaneously recharging the batteries.

For the geeky geeks, here is a graph from a day last week in Pukapuka (Cook Islands); a system that Dean installed December last year. The blue line is the village load. Red line is the solar energy utilised by the system. Light blue line indicates what potential solar energy was available to harness on that day. But batteries were already charged by 11am and the system was clever enough to only draw in what was needed for the rest of the day to cover village load and keep topping the batteries up ready for the next overnight drain.

And here’s another cool graph.

Worked it out yet?

This is how energy flows through the system. Overnight, energy is being drawn from the system at a relatively constant rate and the battery state of charge slowly drops. When the sun comes up, suddenly there is a lot more energy coming into the system than going out. The batteries are charged by 11am and the solar inverters slow down the draw from the array. Throughout the remainder of the day the system draws enough energy to power the village and keep topping up the batteries. Then the sun goes down and it all happens again.

Note: each line represents a different cluster of batteries and their different rate of charge. SMA charge algorithms were charging the batteries at different rates this day.

So that was class #1. Still following? Time for class #2.

I said before that we are harnessing energy using component building blocks which allow us to package it up into the most efficiently transportable resource so that we can distribute it where it is needed or store it up to use later.

Power = voltage x current – these are our electrical basics.

Simply put, voltage is like pressure and current like flow. You can have a skinny pipe with high pressure but low flow sending the same volume of water down the line as a fat pipe with low pressure. The same applies with electricity cables. The voltage at which you choose to move electricity around has a lot to do with the cost and practicality of cable thickness.

Even working with relatively thin cables is hard work

When solar panels are connected in series, the voltage of each panel is summed together. If each panel has is rated to 34 Volts and 8 Amps, a string of 10 panels connected in series will deliver 340 Volts at 8 Amps. Connect two of these strings in parallel and you have 340 Volts at 16 Amps.

At the front end of the system on Vaitupu is the solar array comprising 1,608 solar panels. Each solar panel has a voltage of 38V and can produce a current of 8 amps at full power. The electrical configuration of the array comprises strings of 24 panels connected in series delivering 912V-open circuit voltage at 8A. These are then paralleled going into the inverter to raise the current to 32A resulting in a total solar inverter capacity of 410kW across 15 inverters.

The solar panels are connected to solar inverters. These inverters change the direct current (DC) energy from the solar array to alternating (AC) energy that can be used to power homes.

At the other end of the system is the battery bank comprising 576 2V batteries. 24 of these batteries are connected in series to produce the equivalent of a single 48V battery with a current capacity of 3,670 Ah. Two of these banks are then connected in parallel to produce a 48V 7,040 Ah battery cluster weighing just under 10T. Vaitupu has 12 of these battery clusters and is the biggest system we will build in Tuvalu.

This is what 120 tonnes of electrical storage looks like

The batteries are connected to 36 battery inverters. These inverters change the DC energy stored in the batteries into AC energy that can be used to power homes. These inverters have the technology move energy bi-directionally. They can charge the batteries when the sun is shining, or draw energy from the batteries at night to power the village.

Everything is connected to the multi-cluster box, which runs the show.

The whole system is based on series and parallel. The solar panels are connected in series and parallel to raise both the voltage and current, the solar inverters are connected in parallel. The batteries are connected in series, and then in parallel. The battery inverters are connected in parallel.

Why? Series allows us to build the voltage up high enough that transmission of the energy becomes practical; we would need a cable as thick as my leg and as heavy as a tractor to transmit 38V (and very high current) electricity from the array. And having many of the same unit running in parallel creates redundancy in the system. This redundancy means that in the event of a failure in any part of the system the system as a whole will continue to operate. It also means that this section of the system can be shut down and isolated for maintenance without having to turn the power off completely.

Just like Lego after all.

On Friday this week we have the 6^th and 7^th form (Year 11 and 12) physics students visiting the site. We’re excited to open the doors and let people see what is being built here. Hadley will be fielding the hard questions.