Michael goes off grid for power: part 1

The sun has powered electricity in my three bedroom inner Sydney house since March 2105.

It’s been six months since the wires from the poles carrying mains electricity in the street outside my house were disconnected.

Now, with useful data on costs and energy use, it’s possible to pass on the lessons I’ve learnt.  You can use the data to go off grid yourself and avoid my mistakes.  The data enables you to compare my experience with yours, and to compare the fifteen or so other off-grid battery and inverter products on the Australian market, with more coming.

Before we get to the energy data here some lessons I’ve learnt (data and discussion on costs will come in the next blog post in two weeks – Michael goes off-grid for power: part 2):


1. Get your baseload as low as possible – it's the key to success

“Baseload” is the amount of electricity the house uses when everything is, or is supposed to be, turned off.  It will be the lights on your clock, stove, computer, hi-fi, wiFi, and your fridge and the battery system itself.

The lower the baseload the longer your batteries last, and the more effective is your dollar investment in the panels and batteries.

To measure your baseload you can; install an energy monitor (eg Wattwatchers, Efergy, Belkin Wemo), or read your energy meter just before going to bed at night and just before you start the day with a shower.

If, say, your baseload is an average of 300 watts overnight then that will be the load through the day plus the load of your shower, toaster, oven, computer and other things you turn on.  A baseload of 300 watts over 24 hours mounts up to a total of 7.2 kWh.  If during the day, you use an additional 5kWh, then your total daily average use is over 12 kWh. 

The cost of solar panels, inverter and batteries to go off-grid for a daily average use of 12 kWh is presently over $18,000 and more typically around $30,000.  If the total daily average use is around 4 kWh there are off-grid systems for no more than $18,000 but, if you only need an inverter and batteries then for less than $10,000.

2. Use most electricity when the sun shines

This seems obvious but it’s taken me a few months to change my pre-disconnection habits of putting the washing machine on early in the morning and using the hose on my garden (which is powered by a pump as the house is not connected to mains water).  Early morning mains electricity is roughly half as cheap as power bought after 12 or 2 or so – it varies from state to state.  With the sun at its most powerful in the middle of the day there’s either more energy than the batteries can store or there’s time to recharge the batteries after the appliances finish running.

3. Carefully choose your system: do you want to waste excess sun energy?

My system (Alpha-Ess) rejects surplus solar energy when the batteries are full.  Not good.

Other system controllers will take solar energy and run it directly to the fridge or other appliances and bypass the batteries, reducing losses and not shorten their running life. One such inverter is made by SMA – the Sunny Island.

4. Expect your suppliers to be learning, too – be patient during the process

The lesson here is, this is new technology for both the buyers and the installers.  To get the product and service we may legitimately expect it’s essential, in my experience, to be specific, clear and curious about what you want, what’s promised, what’s happening.  Treat hiccups as learning experiences and, so long as there’s good will both ways, you can get what you buy.

A strange thing, your engineer.  Well, most of them – there are wonderful and growing numbers of exceptions. 

Brought up to wonder at the curious and open mind and imagination and communication skills of engineers such as Archimedes, Leonardo Da Vinci, Michelangelo it’s been a life-long disappointment for me to endure the communication-resistant ways of many of today’s engineers. They speak a different, often obscure language and the idea of explaining and discussing ideas – of giving a service - with customers is bizarre to them.  I blame universities (including the one where I teach in the engineering faculty) who don't teach communication skills. 

A little example.  After the new solar panels were put up the gas booster for the hot water system, which is located among the panels, there was a sustained period of rain and the booster wouldn’t work.  Responsibility was denied by the panel installers.  Upon inspection (at a cost of $220) of the roof and system by the Solahart hot water plumber / electrician it was discovered the gas tap on the roof had been turned off, either to allow access to fit a new panel or to make it safe to work with electrical drills and any sparks, and the wiring to the gas ignition had been pulled loose, perhaps by an accidentally placed boot. This is no big deal but it displays the tendency of most engineers to deny first and investigate or accept responsibility second - if compelled.

In the next blog, Part 2, I’ll detail the costs and, if answers have been received to the issues above then I’ll pass that on, too.




In this section, I’ll be sharing production and consumption data reported by four different systems and thought it helpful to give an overview of what these systems are, where they are located in my house, and what they measure.

  • Tigo energy monitoring: a solar optimiser that provides module-level monitoring; it is located on the roof and measures the amount of solar energy produced by the panels
  • Alpha ESS: energy storage system that holds inverter controller and lithium batteries; it is located between the meter box and the solar panels, and measures the solar energy it receives (noted in this article as production) and the amount of energy consumed by both the house and the system itself.
  • Efergy energy monitoring: a WiFi-based energy monitoring system; it is located inside the meter box and measures the energy consumed by the house.
  • Wattwatchers energy monitoring: a WiFi-based energy monitoring system; it is located inside the meter box and measures the energy consumed by the house (similar to Efergy).

Diagram of the monitoring systems and their locations:



As I mentioned above in “lessons learned”, getting your baseload as low as possible is the key to succeeding in off-grid living and maximising your solar and battery investment. It was very important that I determine what my base load was, identify what appliances or systems were contributing to it, and see if there were ways I could lower it. I was confident I knew the base load of my house, with my wastewater aerator and refrigerator being the two main contributors, but I was very curious to find out the energy draw for the Alpha ESS system.

Step one, determining my total base load, wasn’t as easy as I expected, especially given the fact that I have three different monitoring systems that could provide me with the information. The Efergy and Wattwatchers systems confirmed what I already knew: my house’s base load was about 86W (60W for the aerator and roughly 20W for the fridge occasionally turning on). However, where I ran into problems was with the Alpha ESS reporting system: it was saying my base load was 257W, which is three times larger than the base load reported for the house. At first I thought this difference of 171W was the base load of the Alpha system itself, but their numbers just didn’t add up.

If you take only the base load that Alpha was reporting of 257W and extrapolate that throughout the day, you get a total daily consumption of over 6kWh, which is higher than the average 5.33kWh they were reporting! Something is clearly wrong with the reported numbers, and I suspect it is a software or calibration issue. I tried getting answers from the installer, but at the time of this article’s publication I still hadn’t received any information that justified this discrepancy.

When I compared the average, daily consumption values reported by the three systems, Efergy and Wattwatchers again were pretty similar, and Alpha was reporting a consumption that was roughly 1kWh higher than the other two. If this additional 1kWh reported by Alpha is accurate, then the system’s base load is actually closer to 40W (very different from the 171W it’s reporting).

Although 40W may not seem like much, for my house this is significant and over the course of the day represents about 20% of the total energy I consume.

On top of that, the system specs claim a <2W consumption. Again, something seems off, but if Alpha is indeed taking a constant draw of 40W, that’s a steep energy price to pay to manage my solar power!

The table below summarises the base load and daily consumption values reported by the monitoring systems.

To sum it up, here are my unanswered questions about my energy consumption:

  • What is my base load and daily energy consumption including the energy used to run the Alpha ESS system?
  • Why are the base load and daily consumption values for Alpha contradictory? Is this a software or calibration issue?
  • Is there a way to reduce the energy consumption by the Alpha system? Is the system’s current consumption reasonable compared to other products on the market?


Since my house first began producing solar power in 1996, there have been two major upgrades. Below is a summary of the three installations and the total system capacity after each one:

  • 1996: 18 x 120W panels installed = 2.16kW total capacity
  • March 2015: 6 x 327W panels installed, 6 x 120W panels removed = 3.40kW total capacitY
  • August 2015: 6 x 260W panels installed, 9 x 120W panels removed = 3.52kW total capacity

I have also included a table with the expected production values for my panels at each stage. The values are based on a chart provided by Australia Wide Solar that was scaled the match my system’s current capacity. Note that these are standard averages over time and can’t be used as precise comparisons to what my panels are actually producing, but I still find the table useful as a general benchmark to understand how much energy my panels should be giving me throughout the year.

Because I continued to experience system outages even after installing additional solar panels and batteries in August, I chose a period of six days in August when I experienced no outages in order to evaluate my daily solar production. The results are shown in the graph below:

As summarised in the table, production for the month of August should be roughly 10kWh per day. The highest production day in my sample week was 8.1kWh. Given variations in weather and the short sample period, this 2kWh difference isn’t really a concern. Once I have longer periods without outages and I can analyse larger datasets, I will be able to more confidently compare my panels’ energy production to the expected daily values.

What does interest and slightly concern me about the graph however, is the 1kWh difference between the reported production values by Alpha and Tigo. Again, understanding what these two devices are measuring is important: Tigo is measuring solar production at the panels while Alpha is measuring the solar energy the system receives. So, (assuming the Alpha values are correct and don’t reflect a software or calibration issue) it would seem that the 1kWh difference is the loss in energy between the panels and the Alpha system, and this (like the discrepancy in consumption) is pretty high.

Like any system that transfers and converts energy from one form to another, there are going to be losses. No system is perfect. However, as I started doing more research, I became aware of a key element of the way the Alpha system operates that may mean my decision to purchase it was a huge mistake: the Alpha system transfers all its incoming solar energy through the batteries before it delivers it to the house. When I learned this, I was devastated. One of the most important figures of merit in a system such as mine are the battery losses. If you put 1kWh into a battery it doesn’t all come out! There are losses associated with both charging and discharging. The higher the charge/discharge rate, the greater proportion of energy is lost and the shorter my battery life becomes. So, I repeat, all my energy is getting charged and discharged through the batteries before I ever even see it in the house. For someone living off-grid, this level of energy loss and battery depreciation is unacceptable, and I was never made aware of it by the installer.



Another huge flaw with the Alpha system that I’ve recently become aware of also stems from the fact that all the energy first goes through the batteries: the Alpha system’s output is always limited to 3,000W regardless of the solar size; it can’t deliver above this. This is an extremely important point to understand because it affects the way I live and how I’m able to use my appliances. I’ll break it down in a way that’s practical and simple; prepare yourself to be blown away by this outrageous system limitation.

We’ve already established that the base load of my house is 86W. Let’s say I wake up in the morning, turn on a couple of lights in the kitchen because it’s still dark (20W), turn on the toaster because I’m in the mood for toast with butter for breakfast (1,200W), and my daughter (who happens to be staying with me) turns on her hair dryer while getting ready (1,500W) and she decides she needs to put on a load of laundry before she leaves the house (500W). Doesn’t seem too out of the ordinary, right? Well, we would be in trouble: all of the power would cut off, and the Alpha system would shut down because we would have exceeded its 3,000W limit. Regardless of the size of my solar system, I can NEVER exceed 3,000W of power consumption in my house while using the Alpha system. This was very hard to swallow.



Now you see why I said that I probably made a huge mistake by purchasing the Alpha system when going off-grid. The simple truth is that the Alpha system is not designed to be used in an off-grid setting, and they have not implemented the necessary retrofits to make it work in that environment. However, during my recent research, I came across a product that is designed specifically to be used off-grid and shows great promise for high efficiency and effective energy management: the SMA Sunny Island system.

The main difference between the Alpha and Sunny Island system: Sunny Island can send solar energy directly to the house when it is needed and completely bypasses the system’s batteries. SMA’s Sunny Island system not only extends battery life by not cycling all loads through them, but using solar directly into loads means items can be set to run on timers during the day, (washing, dishwasher etc) to maximise the benefit of an abundant afternoon supply of solar. It also has a larger peak design capacity than Alpha. For example, if you have a 4kW solar system, with the SMA units that would allow a potential delivery of 4kW of solar (in optimum conditions) directly into the house’s load + the 4.6kWof power from the batteries delivered by the Sunny Island controller (they can run in parellel to each other).  That’s a big potential 8.6 kW of continuous capacity to loads.  As I’ve already pointed out, in contrast the Alpha output is always limited to the 3,000W delivery of the battery inverter regardless of the solar size.

Below is a diagram to help visualise the differences between the way the two systems operate and manage solar energy:

To sum it up, here’s what I’ve learned about my Alpha storage system:

  • The consumption and production values that Alpha is reporting are unreliable, and the system seems to not be calibrated correctly or is having software issues
  • Alpha has an inefficient way of managing my solar energy (by diverting all of it through my batteries first), which
    • decreases my battery life by constantly charging and discharging them
    • increases the system’s energy losses
    • caps my power usage at 3,000W (which we saw is not hard to reach)
  • There are other options out there (such as SMA’s Sunny Island system) that maximise solar production and battery performance by effectively managing the energy and where it goes