Renewable Energy Solar Energy Wind Energy Geothermal Energy Bioenergy Hydropower

Time for a Change: Micro-inverters Improve Performance of Solar Systems

By Raghu Belur, Enphase Energy
September 8, 2008 | 27 Comments

Do you like this opinion & commentary?

Bookmark

Feed

There are many challenges in current solar technology. These challenges include the inefficiencies in photon-to-electron conversion, system reliability, and the difficulties in installing and managing solar installations. The majority of current research and development funding is looking to address the first issue by focusing on improvements in solar module technology and manufacturing. While this is important to drive down the cost of solar modules, there are additional opportunities for improvement in solar system performance. Re-thinking the role inverters can play in solar installations is one way to address the challenges associated with energy harvest, reliability and management of solar systems.

Traditional Inverters

Inverters perform two key functions — converting the DC power from the solar modules into grid quality AC power and performing Maximum Power Point Tracking (MPPT) on the solar array. MPPT is the algorithm that extracts the maximum amount of power from the solar array.

Traditional centralized inverters manage all the solar modules in an array as an aggregate — a single source of energy. But there are limitations in using a single MPPT for multiple solar modules. Non-uniform changes in temperature, irradiance and shading create complex heterogeneous current-voltage curves, making it difficult for an MPPT algorithm to operate efficiently. This can result in the conversion of less than the maximum power available from the solar modules. An associated issue is that a traditional inverter represents a single point of failure in the solar system. A final limitation of traditional inverters includes noise pollution, space constraints and aesthetic issues. The larger the solar power system, the larger the inverter that is required, sometimes requiring a separate facility that must be constructed, powered, cooled and maintained.

The Micro-inverter Systems Approach

The idea of a micro-inverter has existed for many years — using multiple small inverters instead of a large, centralized inverter to distribute the power conversion of a solar installation. Early designs met with limited success because they failed to achieve the efficiency, reliability and price point required to achieve regulatory and commercial viability. More recent designs such as one from from Enphase have taken these lessons to heart, and micro-inverters are being developed and tested to meet the industry regulations and customer requirements. What is emerging is a micro-inverter that produces more energy, is more reliable and less expensive to install and maintain than traditional inverters.

Harvest Gains, Reliability, Design and Installation

Micro-inverter technology offers significant technological advances. The first is per-module MPPT, which ensures energy harvest is maximized. With traditional inverters, the performance of the entire array is degraded if one or more modules becomes dusty, covered in debris or shaded. Micro-inverters enable each module to perform independently within the solar array. This benefits the system owner by maximizing energy harvest since degraded performance from one solar module will not prevent the rest of the modules from producing their maximum energy. The per-module micro-inverter also eliminates the problem of reduced energy harvest due to module mismatch. Tests have shown an increase in energy harvest in the order of 5 to 25%.

As with the migration from a central mainframe to personal computers in the information technology industry, the distributed nature of the micro-inverter system significantly alters the landscape of the solar industry. A centralized inverter failure in a traditional design renders the entire solar array useless until a replacement can be ordered and installed. With a micro-inverter approach, a single inverter failure only affects the module to which it's attached, and contributes to an insignificant degradation in total system output. This means that micro-inverters can be replaced at the convenience of the installer or during a regularly scheduled maintenance visit. In addition, no specialized personnel, equipment or tools are required to replace a micro-inverter. This distributed approach can achieve system availability of greater than 99.8 percent — a key factor for commercial systems.

The micro-inverter approach also promises to increase design flexibility and lower installation costs. Micro-inverters are connected in AC branch circuits eliminating the need for expensive DC combiner boxes, disconnects etc. This means that installers no longer have to deal with restrictions of string design or with the headache of locating and installing a large traditional inverter. Individual solar modules no longer have to be "matched" because each module is an independent energy producer with its own MPPT. Use of micro-inverters eliminates the hassle of working around varying module ratings and standardizes the installation process. The per-module inverter approach means that installers have the flexibility to install solar modules on any available roof surface without the restriction of coplanarity.

Enphase Energy has extended the concept of a micro-inverter into a complete system, which includes the micro-inverter, embedded powerline communications and Enlighten — a web-based performance monitoring and analysis tool. The tool is able to compare modules and if it notices an energy harvest issue, it uses built-in algorithms to analyze the cause of the issue and whether the owner or installer needs to be notified.

With a systems approach to solar generation that delivers greater productivity, higher reliability and smart management, micro-inverter systems significantly improve the performance of solar systems.

Raghu Belur is a co-founder and vice president of marketing at Enphase Energy Inc. He has 15 years of experience in engineering and engineering management. He began his career at the Indian Institute of Science where he worked on a team developing an alternative energy gasification system. He was an early engineer at Cerent Corporation where he was responsible for the development of the optical 2.5Gig interface for the Cerent-454 ADM. Cerent was acquired by Cisco Systems in 1999. At Cisco, Raghu was managing the team developing the 10Gig interface products. He co-founded Enphase Energy Inc in March of 2006. Raghu has a MSEE from Texas A&M University.

Solar Energy

The information and views expressed in this article are those of the author and not necessarily those of RenewableEnergyWorld.com or the companies that advertise on its Web site and other publications.

27 Reader Comments

Comment
1 of 27

deep_patel

September 8, 2008

I like the concept of the Mirco Inverters, but it seems like on the residential and light commercial side, these inverters do not make much sense because you have to use a monocrystalline panel and the financial aspect. Larger generation projects may benefit from this inverter. Have you conducted field tests or just lab tests?

Comment
2 of 27

Jonathan_Cole

September 8, 2008

Excellent, Raghu! I have been trying to promote this approach for decades. Finally it is getting done. Another incredible advantage to this approach is longevity. Small inverters (and charge controllers) at 90%+ efficiency dissipate hardly any heat because the load of the entire array is distributed into such small packets. This means the long-term heating issues that degrade components in larger inverters are no longer an issue. It means that passive heat dissipation such as radiative cooling and micro heat pumps are viable instead of having to rely on fan driven systems which are a point of failure in a system made to last 15 years or more. This amounts to longer life and lower cost.

Since such mini inverter modules can be plugged into the back of a solar panel, much as you might plug a nightlight into a receptacle, replacement takes minutes. This then means that advanced technology upgrades are viable which is important over the life of a solar electric system since it may be as log as 40 or 50 years.

I have a friend who has been operating a 300 watt inverter continuously for 15 years without a failure. Packaging all components into individual panels is the way to go. Now when we include energy storage such as the lithium titanate batteries being offered by Altairnano, we come up with the ultimate self charging uninterruptible power supply. Take it to mass production, and the energy problem is on its way to being solved.
http://lightontheearth.blogspot.com/

Comment
3 of 27

robert-lada-81349

September 8, 2008

Yes I see the light, Hah. Truth be told I would invest in a simple 35kw system as discribed in this article .. Great Job!!! I know a few people who would purchase such a system.. NATCH!

Comment
4 of 27

ben-gatti-37189

September 8, 2008

Jonathan,
Spreading the heat components around, makes it more expensive to manage the heat not less. When the heat is close, you can use fans, heat-sinks, and temperature sensors to control heat, with active heat control, one can use the full capacity of the components.

If many large inverters where more cost effective than larger ones, there would be no market and a negative price curve for larger inverters. In Fact, inverters at the high end of the power continuum are about half the cost per watt as those at the low end.

Having serviced residential and commercial low-voltage systems in the past, I am somewhat skeptical of the savings long-term. A single point of failure can be very easy to diagnose; while finding a problem somewhere on a high, hot, and sloped roof, at the end of a long ladder, is more dangerous and more expensive.

On flat commercial roofs, you can get away with equipment spaghetti, but on a sloped roof; I would prefer as little as possible up there, and as much as possible mounted in a cool place at eye level, but that's the been-there-done-that part of me talking.

I do like the idea of better MPPT efficiency, and fully scalable systems; But micro-inverters will need to be fully integrated into the Panel before the pros outweigh the cons.

Comment
5 of 27

Jonathan_Cole

September 9, 2008

Ben, You clearly have not looked at the instrumentation demo on EnPhase's website. With this advance in instrumentation there is no longer any diagnosis necessary. You know what you need too fix before you even go up on the roof. And while economies of scale may now favor large inverters, that is only because all of these products are not yet in the true realm of advanced automated mass production. When these things are being manufactured in the hundreds of millions of units, I guarantee you that they will be dirt cheap.

Comment
6 of 27

ben-gatti-37189

September 9, 2008

Jonathan,
When I was repairing DC systems, we also had diagnostic panels; this doesn't always mean you know where the problem is.
In a single inverter plan, one may not need to go on the roof to fix the problem.

Your second point re: Automation - I don't know - how do you automate making all those extra connections?

Standard Panel = 2 connections per panel
Micro-inverter = 2 + 2 connections per panel (Plus data?)

So at least twice as many electric connections.

What is the basis of this "dirt cheap"?
Each unit requires a microprocessor, some high voltage switches, hi and low voltage supplies, some data transmission. Any larger system would have the same components with the sole addition of larger switches.

It is really hard to imagine how smaller is cheaper-per-watt.

Ben

Comment
7 of 27

ben-gatti-37189

September 9, 2008

Back to your last point about adding batteries:

Tell me Jonathan,
how would you add batteries to a Micro-inverter system?

Comment
8 of 27

Jonathan_Cole

September 9, 2008

Ben, I think you may be misinterpreting the way such a system would work. It is true that with a single inverter you would not have to go up on the roof. On the other hand with the micro-inverter system, with each panel putting out AC and running in parallel, if one fails, you won't need to go on the roof until a convenient time. Because even if 15% of your panels fail, you are not out of business. But with a single inverter, failure is a complete power outage. Since small inverters will be more reliable because they don't have mechanical heat dissipation, there should be much less need for repair. The ideal is to have inverters that can last the life of the PV and even to be economically upgradable, since electronic technology continues to improve over the 30-50 year life of the PV.

When the inverter and charge controller are integrated with the panels it means that a single AC wire can come from the entire array which seems like much fewer connections and much less wire cost to me. And the data for the instrumentation can be multiplexed on the electrical signal so that no additional wiring is required for complete AC/DC instrumentation.

As far as connecting batteries into such a system, they are connected between the charge controller and the inverter, just as in every un-integrated independent system - like the one I have been living on for 26 years. I am sure that EnPhase is probably targeting the grid-tied market, but it is only a relatively small step to integrate storage when the economic issues are supportive.

Cheaper per watt has to include the maintenance and installation costs, including wiring, and enclosures for weather protection, etc. Even in Hawaii we have these costs and they are significant. We need an integrated product that can be installed in a morning and supply sophisticated data reports, querying, alarms and instrumentation to have user-friendly, 30-50 year installations with high reliability.

This awaits visionary investors.

Comment
9 of 27

ben-gatti-37189

September 9, 2008

Jonathan,
I appreciate your points - I do, But I think you've failed to make the case for three key points:

1. That a transistor sitting on a roof will run cooler than in the moderate temperatures of a shaded garage. Try this experiment, park your car in a garage, then park it in the hot sun; use the vinyl seat test or some other scientific method. Trust me that roof is a hotter environment.

2. A Single failure won't cause a problem. This I think is a maybe. Ever had a Mosfet fail? It shorts. Not sure what would happen if a few Mosfets short over; maybe not much. But saying doesn't make it so.

3. How do you include batteries in a one-panel, one inverter system? If I understand you correctly, You'd have batteries behind each panel, connected between the "Charge Controller" and the inverter. That would mean climbing the roof to replace the batteries. Now you have to admit that's a bit of a tricky wicket.

Comment
10 of 27

Windmillman

September 10, 2008

I think this is a concept that is so new and innovative that it is a bit hard to grasp. Micro-inverters are a new concept and will be used in different ways. Efficiencies will improve. Systems will become simpler and provide additional benefits. Small moderately expensive inverter systems today in little boxes can become tiny components with a high efficiency and small cost one day. They are being used on solar systems now, but they will be used on applications difficult for us to imagine today. It's like comparing vacuum tubes to transistors. The first transistors came in little metal cans, but many people continued to use vacuum tubes for a while. Now transistors are used at the core of most chips and these are everywhere. If Enphase works hard to continue to improve performance, reduce size and expand applications, their future is unlimited and I wish them great success for their vision for the future. I already have some great ideas how they can be used for other applications. My head is already spinning, just thinking about the possibilities.

Tom Conlon

Comment
11 of 27

allan-sindelar-20506

September 10, 2008

The microinverter approach requires rethinking how we installers approach grid-tie design. This is much like in 2001, when the high-voltage string inverter approach, pioneered by SMA in the U.S., replaced the 48V standard used by Trace SunTie and AEI GC-1000 up to that time. Who would now use a central inverter with multiple 48V strings?

Some of us will embrace this new approach; some of us, such as the ever-negative Mr. Gatti, will resist it. Best to try this approach and share our results openly; we in the industry are on a roller coaster, and the ride will only get wilder.

On a more technical note, I would like to see Enphase offer a 120V AC model of their microinverter (AC output of the current Enphase models is 240V AC), in order to serve a niche market: conversion of existing off-grid systems to grid-tie. We have seen several of our own off-grid installations, typically 8-12 years old, become functional orphans. This typically occurs when a residence gets a new owner and the grid is extended. Existing DC array input has 3 conductors: PV+, PV-, and ground. Many older arrays of 12V modules can't be rewired to produce high-enough DC voltage for most string inverters. The Enphase offers a simple solution, but needs 4 conductors: two AC hots, a reference/communication neutral, and ground. A 120V AC model would allow the existing DC input feed to be converted to carry microinverter AC from the array to the grid connection.

Comment
12 of 27

joel-davidson-60793

September 10, 2008

I am not convinced that micro-inverters or even multiple small inverters are the way to go for many PV installations. I installed the first OK4U inverters in the US and did a side-by-side 400 kW Xantrex vs. SunnyBoy system so I know micro-inverters and multiple string inverters can work. However, 75 to 85 C module temperatures and extreme daily rooftop temperature swings are a tough environment for electronics. Running AC into a branch circuit is simple, but you still have to meet local utility disconnect requirements. Redundancy is sometimes good, but fewer moving and non-moving parts are better. I hope Enphase the best of success, but also would like them to prove their inverter's reliability under extreme conditions, show cost comparisons as Multi-Contact did with j-boxes, and join the PV industry battle to end utility company required AC disconnects.

Comment
13 of 27

ben-gatti-37189

September 10, 2008

Thomas, Allen,
First, I'm obviously not alone in questioning the core assumptions of Micro-Inverters.

Second, I can see some cases where they may work well, starting with temporary emergency systems (military and relief organizations) which need to be brought up quickly, and scaled ad hoc.

That said, Micro-inverters are simply not compatible with battery-back up - which almost immediately precludes the self-same markets.

Here's a rule of thumb, the number of inverters should match the number of battery-racks or cabinets.

Comment
14 of 27

sunovitch-glenn-126777

September 11, 2008

I purchased and installed 3 Microsines, which are small sine wave inverters designed and built in Holland and sold through Trace Engineering (now Xantrex). All 3 worked fine for 1 month to 6 months before ceasing operation. They are sealed with epoxy resin and can't be investigated for cause of failure, nor can they be repaired. Therefore, I lost my $1000 investment.

Hopefully, the Enphase inverter is both hardier and more easily repairable. I can just imagine a homeowner who, 10 years after installing a 4kW PV system finds out that he's only outputting 1.5kW because 9 of his inverters are dead.

The success or failure of such a product is dependent on design, build quality, life-time testing under extreme environmental and electrical conditions and the quality of the warranty. I wish Mr. Belur only success.

Comment
15 of 27

murray-rose-67503

September 12, 2008

Regarding Joels' -" Utility company required AC disconnects"? Unbelievable, the earth is starting to fry and these clowns are still allowed to play 'corporate cowboys' and hinder clean energy initiatives.

Comment
16 of 27

belur-keshavram-157100

September 12, 2008

Dear Raghu,
All the best .Keep it up. Keep the Indian TriColour Flag Flying
High.Great Idea.
G-Getting
R-Really
E-Exicited
A-About
T-Tomorrow

I-Interesting
D-Decisions
E-Emerges
A-All

W W R
Keshavram

Comment
17 of 27

henk-daalder-156480

September 12, 2008

We have done this about 10 years ago. A small inverter per panel.
Then the producer and the owners of the panels discovered that elekctronics may have a lifetime of only 3 years. A faulty inverter, that is built into a panel,makes the panel useless.
Inverters must be detachable and replaceble.

Furthermore I think that inverter efficiency increases at higher voltages, so some panels in series that feed a single inverter (at a higher voltage than a single panel) works at a higher efficiency then an inverter per panel.

And one replacable inverter per panel means more connections that can fail.

But where only one or a few panels have the same light conditions there a small inverter is essential.
Important is however that the inverter communicates its status and failure mode. Otherwise its failure may be undetected for months or even years, which happened with the dutch "inverter per panel"

Comment
18 of 27

Jonathan_Cole

September 12, 2008

Actually the development of panel mounted inverters already goes back quite a way. The first that I heard of was an Australian firm. They did not have the right design and went under. Lately, Exceltech has come out with this product on the market.
http://www.exeltech.com/pvacproduct.htm

Before you get too sure that this is a bad idea, you might want to look a little closer. Actually underneath a PV panel remains remarkably cool. One quarter inch from the back of my Kyocera panels in Hawaii on a full-sun summer day at noon (we have very strong sun), it is the same as ambient temperature. If you put your hand back there it does not feel any warmer than the air. Most of the heat dissipated from most solar panels rises (because heat rises) and does not come out the back of the panel. This is why here in Hawaii, large hotels are covering there roof with PV, because they generate electricity and lower their cooling costs at the same time. PV on a roof is a heat interceptor and dissipates nearly all of it upward. As far as being able to put a battery pack back there as well, you might want to check out
http://www.b2i.cc/Document/546/NanoSafeBackgrounder060920.pdf

Integrated PV/electronics/storage is the future in my opinion. The fact is, that without storage there is a limit to how much time-variant solar energy can be utilized by the grid. Of course it might put a lot of old style installers out of business, but only if they refuse to see the writing on the wall, because once these products are in mass production, the demand will be enormous. So it might only take a few hours for the installation of the entire system, but there will be millions of systems to install instead of tens of thousands.

Don't forget that 100 years ago, every automobile was a custom, one of a kind integration of a hodgepodge of parts. Mass production is what changed things and it will be no different with integrated PV/electronics/storage systems

Comment
19 of 27

Jonathan_Cole

September 12, 2008

If you think that storage integrated with PV panels is not viable because it would be too difficult to replace the batteries....Altairnano batteries(see link above) have been tested to 15,000 deep cycles with better than 90% in/out efficiency. At one deep cycle per day, that would be 41 years. Supercapacitors have been tested to 600,000 deep cycles, so storage that operates between 30 below zero and 140 above zero and lasts as long as the PV is right around the corner. Mostly it is awaiting the investment to undertake large scale mass production.

Comment
20 of 27

denny-houghton-50945

September 13, 2008

From the utility lineman's point of view. We support the right of the "corporate cowboys" to prevent you from connecting a Non-UL listed inverter to the grid. In fact we support that concept with our lives. Get over being inspected and regulated. This is not harming the renewable energy industry. Uncertified and possibly dangerous equipment installed by unqualified workers may cause much more harm to the industry's credibility in the long run.

Comment
21 of 27

denny-houghton-50945

September 13, 2008

I stand corrected.
Enphase products are CSA approved. CSA is a NRTL.
Anti-islanding capability is critical for line worker safety.

Regulatory and Compliance

Are Enphase devices UL-approved?
The Enphase Micro-inverters are CSA listed per the requirements of UL1741 and are therefore approved to operate in the United States and Canada

Comment
22 of 27

ben-gatti-37189

September 13, 2008

Jonathan,

I fail to see how buggy whip platitudes and this cult mentality of "loosing out for not seeing the light" helps make the case for a particular configuration.

The challenge is absolute cost, and managing the inconsistent lifetimes of various components. If Solar panels last 20 years and the batteries last 40, why would you integrate them? I have a hard time seeing the play here.

Comment
23 of 27

Jonathan_Cole

September 14, 2008

Ben,
I don't see how "buggy whip platitudes" is anything but a non-sequitur by some one who clearly is not very experienced with these systems. Although PV panels have 20 year warranties they have a much longer life. Maybe as long as 40 years. I don't know what you are talking about, "not loosing out for not seeing the light".

I have been living with and creating PV solar systems, concepts and products for 26 years. My perspective on the future of this technology results from a study of the technology, backed up by decades of experience, invention and engineering of working, idiot-proof systems. This experience along with a masters degree in business tells me that integration is the future for PV. Integrated and distributed systems. Its already moving in that direction. Look at the Sharp grid tied inverter that integrates the charge controllers. You still have to find a safe place for that equipment and it still incorporates a high heat output even if it is 90% efficient. A 5000 watt inverter at 90% efficiency has to be instantly capable of dissipating 500 watts of heat energy. If the fan fails the thing shuts down until some guy like you can manage to find time to come and repair it. If we want wide spread adoption of this technology we cannot have a lot of bad experience stories circulating. I know many people who for the failure of an inverter have given up on solar as too much hassle.

We need idiot-proof products. That means do not leave it to a hundred thousand installers designing different systems. In my opinion. You are welcome to yours.

Comment
24 of 27

Jonathan_Cole

September 14, 2008

By the way, many of these issues are addressed in EnPhase's video
at http://www.enphaseenergy.com/video/Enphase_Corporate.mov

119 years mean time between failure sounds acceptable to me. I just wish they would take the next obvious step and instead of offering a 10 year warranty to offer one that matches the PV warranty, even as an extended warranty option.

Comment
25 of 27

ben-gatti-37189

September 15, 2008

Jonathan,
Generally your points are reasonable, but I thought this bit sorta went over the deep end...

"Of course it might put a lot of old style installers out of business, but only if they refuse to see the writing on the wall,... Don't forget that 100 years ago, every automobile was a custom..."

This I characterized as "Buggy whip platitudes" and cult mentality. Perhaps one isn't familiar with cults, which occur in Business as well as religions; but cults are mutually exclusive by definition, and this notion that installers will go out of business if they don't exclude one product over the other. In practice, the smartest installers will dip their toes in the water cautiously, offer both products and let the facts develop and the market decide. This veiled threat that installers will go bankrupt if they don't "see your personal light" is exactly the bullsh^&t one hears from cult leaders.

I expect you didn't intend to come off like a cult leader, but given the accumulation of cult-personalities, and cult-companies such as CitizenRE as clean energy becomes more interesting, a more serious writer might consider avoiding cult-like phrases.

Comment
26 of 27

jszrom

September 16, 2008

As an electrical engineer that works in the building industry, I have a question for some of the more experienced solar installers/contractors out there.

It seems that the installation cost would be the same if not higher for the micro-inverter design for these reasons:
- 2x as many electrical connections to make (input and output of each inverter)
- Mounting of each micro-inverter
- 240V wire not previously needed connected to each module and a "circuit" is still needed for a group of modules (i.e. you can only put 16A on a #12AWG wire). Therefore you still have virtual "strings".
- Conduit not previously needed for the 240V wire connected to every module.
- Depending on the size and location of the array combiner/distribution equipment is still needed to group the "circuits" together.

Other than these installation concerns, I think the system has many positives, especially the power line communication and monitoring aspect.

Comment
27 of 27

chuck-forbes-38371

October 22, 2008

In mid July at about 1PM in northern NJ, I made a service call on a roof in bright sun. There, under the northern most panel less than 4 inches above the roof was a live raccoon, obviously enjoying the shade. He did not seem bothered by high temperatures. I would hope the electronics could perform as well. The raccoon had downgraded an entire array by chewing a wire from a single panel. I hope this makes economic sense. I am contacting my supplier.

Add Your Comment

Registered users, please make sure to Sign-In. We and others want to know your ideas and opinions. If you are not yet Registered -- it's quick and easy. Just click below.
Thanks!