Microinverters Vs. Dc Optimisers: Which option is best?

optimisers and microinverters

If you’ve decided on panel-level optimisation, which is the better choice, microinverters or optimisers?

Good grief this article is going to get me into trouble with some solar installers.  Why? Because by the end of it I will give my humble opinion on whether microinverters (e.g. Enphase) are better than DC optimisers (e.g. SolarEdge or Tigo).

Please let me be very clear. This article is not saying microinverters or DC optimisers are a better choice than a standard central string inverter. That totally depends on your situation.

If you have an unshaded roof, and it is important that your system is battery ready, then a conventional string inverter system is a very compelling choice.

However, if you have a multi-faceted roof and or shading issues, then individual panel optimisation is almost essential. Your choice then becomes whether you achieve this with micro-inverter or DC optimiser technology.

But first – let’s get back to basics and explain the jargon I’ve just hit you with!

(If you prefer graphics to text this post is published as an infographic here)

The 3 ways you can connect solar panels together

 

1. String inverters

The traditional and most commonly used method is to use a single (string) inverter. On the upside it’s low cost, there’s a huge variety of products available and the technology is well understood.

Power is distributed at high voltage DC (typically 200-500VDC) to the inverter where it is converted to 240V AC. On the downside, because all the solar panels are connected together in series, individual panel outputs are not optimised. If shading, soiling or different temperatures occur the entire output can be dramatically reduced or even stopped.

string inverter connected to panels

A conventional system with a string inverter (image: Enecsys)

 

This is where “panel level” optimisation can help.

There are two primary methods used to achieve this that we will explore today– power optimizers (also called DC optimisers) and micro-inverters.

2.  Power Optimizers

Power optimizers essentially split the traditional string inverter into two separate parts:

DC Power Optimisers on a panel array

A solar system using Power Optimisers (image: Enecsys)

The panel optimizer is attached to the solar panel and thus gets the benefits of panel level optimization under a wider variety of conditions. Different manufacturers each claim different advantages and some partner with panel manufacturers to factory fit them.

The power generated is transmitted at high voltage DC (typically 200-500VDC) and the inverter, typically wall mounted, does the DC to AC conversion. Therefore you get the benefits of panel level optimisation and the cost savings of using a single string inverter.

3. Micro-Inverters

Micro inverters simply take all the functions of a string inverter and miniaturise it to solar panel level. Consequently, each solar panel has its own integrated optimizer and inverter attached either on or under each individual solar panel on the roof:

microinverters on a panel array

A microinverter based solar system

The power generated is transmitted at 240V AC from each micro-inverter and connected in parallel, then connected in directly to your switch board. Different manufacturers each claim different advantages and have a few minor twists on the theme, but the principle of operation remains the same.

Pros and Cons: Micros vs Optimisers

 

We have taken a look at the key issues that differentiate micro-inverters against optimizers and listed them below for you. In all cases, we are comparing to the better/high quality products because in our view, low end inverters, no matter what type, are a nightmare to be avoided!

Efficiency

As a general rule string inverters are typically marginally higher in efficiency. However, your overall system efficiency (yield) should be slightly higher with optimizers or microinverters because they optimise individual panel output. The difference in claimed efficiencies between optimizers and micro inverters is so little, it’s barely worth considering.

Draw.

Installation

Optimizers require a single inverter to be installed and DC cabling d protection equipment on the roof. Micro inverters do not and instead, use plug and play connectors to install so from what I hear they are faster. However, if your module already has an optimizer fitted then you are installing single inverter only plus the cabling, this option would be marginally faster. Theoretically, the efficiency of DC or AC transmission from the roof should be about the same, assuming standards are followed.

Draw.

Flexibility & Expansion

Because optimizers need to be designed to match the “maximum panels per string” on string inverters, the design parts takes a little extra time. I also note that not all inverters work with all optimizers, so you need to get some product familiarity training, although there are some neat tools to help you.

Micro inverters on the other hand only need to match the power output of the panel and then are only limited by the capacity of what they call “AC branches”, or to the layman, the power limit of the AC cable and plug system. In most residential applications you won’t hit this limit.

Additionally, if you want to add more panels later, a micro inverter system can easily accommodate this as long as you don’t reach the AC Branch limits, and you can add them one at a time if needed. Optimizers need to be matched to the string inverter limits, so that’s a harder boundary.

Win to micros.

Scalability

When you consider larger, commercial installations the theory is that optimizers have two advantages. Firstly, you can go to larger string inverters which get cheaper as they get bigger. Secondly, three phase or “double optimisers” which can take inputs from two panels offer further scalability benefits.

Putting aside the design and installation complexity (compared to micros) I think optimizers may offer a very slight advantage in terms of capital cost due to scalability.

Win to optimizers.

Battery Readiness

Batteries charge and discharge DC electricity. So batteries should be cheaper and easier to integrate with Optimisers, because they keep the DC wiring all the way down to the central inverter. Certainly the Tesla Powerwall is designed to integrate out-of-the box with Power Optimiser company Solar Edge’s inverters.

An microinverter system runs on AC, so battery integration is always going to need an extra battery inverter to “AC couple” it into the system. This will make battery integration more expensive with microinverter systems.

Win to optimizers.

But there is one caveat: The Enphase AC battery. If your microinverter system uses Enphase micros, then in 2016 sometime you will be able to add Enphase’s AC battery very easily. This is a 1.2kWh battery that can be installed with a couple of wires, and is controlled by the existing microinverter control box. It is essentially a battery with a microinverter. Once the first battery is installed subsequent batteries are plug and play. The only problem is, Enphase have not disclosed the price. If the price is reasonable then Enphase microinverter systems become very battery friendly. If the cost of the AC battery is steep, battery lovers may be better off using the super cheap Tesla Powerwall and a power optimiser based system

Robustness

This is a tricky one. On the one hand the logic of having fewer parts on the roof (just the optimizer and not the inverter electronics) is pretty sound. Intuitively, this would suggest a higher probability of reliability to optimizers. However, with the benefit of hindsight in the market I don’t see evidence of any major differences in failure rates (excluding crappy micro inverters). What we also see from the top micro inverter guys is very high volume manufacturing and intense attention to automotive level quality assurance.

It’s also the case that in a micro based system, if one unit fails, the rest of the system keeps on working because there is no single point of failure. In an optimizer based system, if the (single) string inverter fails, the whole system shuts down and in some cases if a single optimizer fails the string goes down too. Having said that, replacing a single inverter in the garage is easier than climbing on the roof.

Theoretically optimizers should be better, but compared to top end micros and overall, I don’t see it, especially when redundancy benefits are factored in. Optimizers have two potential points of failure– the optimizer and the inverter- compared to micros single point and I think that separates them for me.

Another factor to consider is the reliability of the solar panels that the inverter or optimiser is attached to. Potential Induced Degradation (PID) is a failure mode of solar panels that is caused by high voltages across the panel and may not show up for 5-20 years after installation. Microinverters keep the panels at a much lower DC voltage than optimisers and so the theory is that this will reduce the chance of PID in the panels. PID is one of the more common causes solar panels failing in the medium to long term.

Win to micros.

Safety

Around the world the risks of High Voltage DC continue to be debated. When everything is done perfectly, great gear is used and maintenance is regularly conducted, HV DC is just fine and very efficient. However, the reality is that often isn’t the case and so rooftop safety is a growing issue. HV DC and 240V AC both present electrocution risks, but HV DC is more prone to creating fires when it all turns bad.

The major optimizer manufacturers almost all talk about their roof mounted units being able to recognise fault conditions and having the capability to isolate individual panels in the case of a fault so even though the HV DC cabling is still present they seem to have that in hand. What I don’t like is that you still need DC isolators which are the number one cause or problems in Australian solar systems. I also note that some optimizers require an optional device to enable this functionality in some circumstances, leaving room for error.

Micro inverters don’t let the DC out; it’s converted immediately to 240AC. Generally speaking it’s also true that electricians are more familiar with AC. Theoretically based on of all of the above, the risk is slightly lower with micros I think.

Win to micros.

Price

The general market price of an optimizer system (including the inverter and DC protection gear) is lower than most micro based systems although the gap is closing rapidly. So ignoring everything else, optimizers win today.

However, I don’t think that tells the whole story when it come to the end customer because they want reliability, performance, adaptability and the lowest cost of ownership, although they might not say that. They don’t want hassles and prefer flexibility. The installation cost was a draw but there’s arguably a bit more design and specification time required with optimizers. Flexibility went to micros but scalability to optimizers so let’s call that quid pro quo. Lastly you have to factor in reliability and safety issues, which was a close call but I think went to micros.

So on raw price, a Win to optimizers.

On lifetime, overall price, a Win to Micros.

The Result

 

Based on this assessment and our reviews of as many other assessments we could find, I don’t think the jury is out on whether optimizers or micros offer distinct advantages over each other. They each show some benefits and advantages but on balance they almost match each other.

It therefore becomes a subjective matter so I’ll just give my opinion and leave you to make up your own mind too. I think lifetime cost, flexibility and simplicity make sense so I would give it to micros. But if a Tesla Powerwall was on my wish list, that would tip the balance to optimisers due to the cheap integration. What about you? Let rip in the comments.

 

(Thanks to Nigel @ Solar Business for help with this article)

About Finn Peacock

I'm a Chartered Electrical Engineer, Solar and Energy Efficiency nut, dad, and founder of SolarQuotes.com.au. My last "real job" was working for the CSIRO in their renewable energy division.

Comments

  1. Frank Schrever says:

    I have two new smart meters and they would not connect me for feedin until I consolidated them to one meter. I could not get any sense from the power company so I went to the ombudsman who gave it a lot of time. In the end, they said it has to be a net feedin calculation and requires one meter. In this day and age we can’t add two together. Has anyone else had this problem?

  2. lol!…..Dunno that I’d be using this sort of comparison:- ” and intense attention to automotive level quality assurance.”

    I’ve just spent four days and several hundred dollars bringing up to scratch a relatively low-mileage car (Ford), including valves/tappets, timing, carby-kit, etc. (all the ‘normal’ ~ periodic ~ stuff), and then still had to take it down for a dyno-tune/exhaust-analysis.

    …..and it wasn’t even up on a wet and blowy roof!

  3. ps. I STILL can’t see a better ~ and virtually bomb-proof ~ piece of technology than the ‘standard’ grid-connect inverter, and swear by my much-maligned Aero-Sharp which ‘does it all’ with panache and nary a hiccup since the early problems with failing voltage-suppressors was fixed. (For free, without a haggle and apologies for the inconvenience.)

    The still-perfectly-operating replacement took me about TWO MINUTES to install.

    My only gripe is that ~ despite several requests ~ none of the whizkids here and elsewhere have come up with a way of circumventing the ‘anti-islanding’ loop so that it can be used in a stand-alone situation.

    It performs ALL the tasks the pricey alternatives do, and ~ according to the published figures ~ does so significantly more efficiently.

    AND I already own it.

    (any expert views welcome: happy to pay for your time: jasw-AT-y7mail.com )

  4. G’day Finn,

    Thanks for the timely and interesting article mate!

    I have been researching this for some time, and in fact, the desire to integrate storage has delayed me installing any solar up to this point. Though given Elon Musk’s announcement about Powerwall and the continued reduction in costs of systems now, I am now very seriously considering my options.

    In my opinion, one critical aspect in the analysis should be the likely loss(es) that will ultimately occur in any DC-AC-DC conversions that will invariably occur with a micro-inverter system. When you take this into account then a DC-DC coupling of say a SolarEdge solution is probably a winner, though I don’t discount some of the points you raised above.

    Anyway, with this in mind, my house is wired for 3 phase so I am unsure as to whether the SolarEdge inverter or storage interface module would provide any dynamic load sharing across all three phases or whether there is a need to have additional components to facilitate this. I am not sure if the Enphase solution has a buggery box to facilitate 3 phase load sharing/balancing either.

    My end solution is complicated by the fact that I wish to run in-slab heaters, which on average consume about 80-100Kw per day. I would love to run these directly from solar during the day but I recognise that even a 15Kw system (as much as my north facing roof aspects will accommodate) will not cut the mission during winter months …. though I could use the grid to build up a credit during summer months upon which to draw down at a later date. In addition, I am just now looking at EVs or PHEVs to potentially improve the potential ROI of the overall system also.

    Finally, there is perhaps another option that should be mentioned here, though it is hard to beat optimisation at a per module , but it is the new range of hybrid inverters that are coming onto market. The Fronious Symo Hybrid inverter is a case in point here and it is also a preferred solution for the integration of Tesla Powerwalls I notice. Cheers Peter

  5. The AC battery is fundamentally a bad idea. It is simply a battery with a rectifier to convert the AC to DC and inverter that will concert the battery DC to AC. It will result in at least 5% additional energy loss compared to a solaredge/tesla battery system. Enphase AC battery has zero chance of being even close to the cost of the powerwall unit on a per kilowatt basis. It also adds equipment cost and reliability issues. Tesla selected the solaredge process based on the technical and economic advantages.

  6. As installers we used enphase micros for years before switching to optimizers almost exclusively about 12-18 months ago. Over time we’ve had many micros fail causing downtime and loss of productivity for our customers and requiring time from us to go out and replace them. Not a huge issue, but definitely not ideal. On the other hand, we’ve never had one optimizer fail. This gives optimizes the win for efficiency to me. Is this consistent with other installers’ experience ?
    Also, Finn, in your opinion how does the solaredge SafeDC feature improve the safety of optimizers compared to micros?
    Thanks! Dara

    • Hi Dara,

      That’s interesting, thanks.

      1) Do you know what the failure mode of the micro failures was?

      2) How recently have you had to replace one and what model version was it?

      Finn

  7. Doug Jacquier says:

    Hi, Finn
    Just as we thought we were close to a decision on our new string solar system, one company raised the possibility of micro-inverters, especially given that we have some shade issues on a well-treed block. You’ve said elsewhere on your blog that if the prices are close you’d opt for micros but I’m also conscious of the comments about extreme climate stress as an issue for micros. We are in SA where summer temps of 40+ are increasingly common. Your thoughts please.
    Regards
    Doug

  8. Good article. One caveat. In 10 years, batteries will indeed be affordable for almost anyone to go completely off grid as long as they have space for them. Utilities are fueling batteries as well as they continue to increase thier “grid” fees. Here in san diego I am already paying $45/mo just for being able to use the grid even if I use 0 electricity. In 10 years I am sure that same fee will be between $75-$100 because those executives dont want to take a pay cut so they will need to keep revenue and net high even if it destroys the business later. They dont care because they probably wont be there anyway and could liquidate their stock long before it happens.

    That means that micros are a bad choice even if their own battery is cheap for the simply fact that every time you convert you lose a significant amount of power. For micros you will be coughing up about 5% just for conversions. No such issue with Tigo or Solaredge although I am not a fan of solaredge because of their proprietariness. I am rooting for Tigo but they have some issues to iron out.

  9. I have a few questions for you:

    Can a string inverter system be converted to a micro system and would that affect the feed in tariffs in Victoria even if no extra panels are being added?

    Also if micro’s are individual can they be added (as in more panels) without affecting the feed in tariffs?

    One last thing if being able to add capacity to an existing system and not lose your current feed in tariff by creating a what is essentially a new parallel system is possible (as I understand it) up to a certain kwh limit (usually <10kwh), is it possible to add a micro array as the parallel system, when you have an existing string inverter system already installed, e.g. is it possible to have a hybrid system? If possible would that be a good idea or fraught with bad karma?

    I ask the last question in the view that the extra panels would have to be installed on the east west facing sides as its a small roof and the north facing is already filled up and as I understand it reading your website, micros would be better (more efficient) for the extra panels having to be fitted to the other facings (based on efficiency, redundancy, shading and east west facings having peak times/lower light times which micros have better output for in these situations being individual, instead of series)

    • Ronald Brakels says:

      Adam, generally you can change your inverter without losing a high feed-in tariff, as long as the total capacity of the new inverter or inverters is not larger than the old one. So you could replace a string inverter with microinverters provided the total capacity of all the microinverters was equal to or less than that of the old string inverter.

      If you have an existing string inverter system and add panels using microinverters that counts as adding to the total inverter capacity and if you have a high feed-in tariff that will cause you to lose it.

      If you have a high feed-in tariff and want to keep it you can’t add any additional inverter capacity that is connected to the grid. But it is possible to add solar panels and inverters that aren’t connected to the grid. For example, you could use solar panels to run an off-grid pool filter or hot water system, or whatever.

      If you are not on a high feed-in tariff, or simply don’t care about losing it, then your options on how to increase your solar capacity are more open.

  10. Hi
    I was wondering with the high feed in rates can I install a larger capacity system and export limit to the approved system.
    Therefore can use the additional non feed in power for my home and maximising the feed in to the grid.

    • Ronald Brakels says:

      I’m sorry Allan, but any increase in inverter size will result in the loss of a high feed-in tariff. But depending on where you are it may be possible to increase the number of solar panels you have and keep it. You may be able to increase your panel capacity up to 133% of your current inverter capacity. Unfortunately, you can generally only do this if your inverter meets current standards and it is likely an inverter that is receiving a high feed-in tariff is old enough not to.

  11. I was just wondering, you say if you want the powerwall, it would tip the balance towards optimizers, because of the cheap integration.

    But doesn’t the powerwall come with its own built in inverter to be able to be hooked into AC systems so it can not only be re-charged by solar energy but also by off peak mains energy for load shifting purposes as well?

    In that case it should work perfectly well with micro inverters as well out of the box making it just as a viable option, unless there is something I am missing?

    • Ronald Brakels says:

      Hi Adam, Ronald here. This is an older article, so it is referring to the original Powerwall which didn’t have a built in inverter.

Speak Your Mind

*