Tuesday, November 17, 2015

A step-by-step guide to breakthroughs

It’s often said that breakthroughs cannot be scheduled.  I’m here to tell you that this is 20th century thinking.  The statement assumes that the word “breakthrough” is unambiguously defined.   This blog post questions that assumption and provides a step-by-step guide to achieving a breakthrough.  My focus is on batteries, but with a few tweaks, one could adapt this for other areas also. 

Let me begin by saying that in the last century there was a feeling that a breakthrough was thought to be when, for example, you discover a new material for a battery that has, say, higher energy or is safer or something.  Those sorts of breakthroughs then go through the traditional rigmarole of publications, licensing, technology transfer, peer appreciation, awards, products and the rest of the boring stuff that takes 20 years to get settled.  In the age of Twitter, Facebook, Uber, and, Snapchat, this kind of time frame is for the folks unwilling to look to new ways of achieving breakthroughs.  If you belong to this “old” club, I suggest you move on.  This guide will be of no use to you.  

But what is a breakthrough anyway?  As far as I know there is no body that proclaims something a breakthrough (a Pope for science?).  And is there really such a thing as an eureka moment?  Even if you have one, it will be a year before you can reproduce the experiment and get all the techniques in place to prove it.  And if the breakthrough is supposed to be a product, it will take you 10 more years to scale it and make it.

But what if there were a reputable publication that actually called something a breakthrough.  And then this was validated and verified by other publications saying the same thing?  That appears to be in line with the scientific method, does it not?

So, for the purposes of moving forward, let us define a breakthrough as just that: It is proclaimed as such by more than one publication.  Also to help us move forward, publications will be broadly classified as a peer reviewed journal article, or a newspaper, or a blog, or a tweet etc.  i.e., as long as the word breakthrough and your work appear on the World Wide Web somewhere, you are golden.  This guide will help you get there.

A disclaimer:  The results are only guaranteed if you follow each and every step.

Step 1:  Before you begin the research, try not to read the literature.  The peer-reviewed literature is full of things that have been tried before.  If you read them carefully, then what you are doing will not be new.  Remember this mantra (courtesy of NBC when they were promoting reruns in the 90s):  “If you haven’t seen it before, its new to you”. 

Step 2:  As you start the research, remember that facts just get in the way.  The literature is full of facts (hence Step 1).   In 1492 everyone thought the world was flat; until Columbus took to the seas.1  Then we all thought it was round, until Tom Friedman proved it was flat.  Until The Matrix came out, we thought gravity was forever binding us to the earth.  Breakthroughs happen when these laws are broken and it takes a bold person to go where no person has gone before.  To paraphrase Marsellus Wallace from Pulp Fiction, you may feel a slight sting every once in a while when it seems like you are violating faraday’s law.  Those are the facts f*ing with you.   f* facts.

Step 3:  Now that you have done your due diligence and ignored everything, it is time to focus.  Try to work on a newly-discovered material, or atleast one that has been forgotten for a while.  This is an important step.  As much as you can go after Steps 1 and 2, the more studied the material, the harder it is to prove to yourself that you are violating all the well-known laws because you are charting a new path rather than screwing up.  It’s so much easier to believe this if it’s a brand new material.  Graphene is good (not as a battery material, but remember Step 2). So are fullerenes  (granted they are old, but it seems like its time to revisit them).  Graphite, on the other hand, could be bad; unless you plan to use it in a new way; in which case it can be good.  Lithium metal is ALWAYS good; but if you go this route you really need to get religion on Steps 1 and 2.  

Step 4:  As you start getting data on the new invention, revisit Step 2.  Revisit it often, especially when you feel down. 

Step 5:  Time to start writing the paper.  Always state that your invention is better than Li-ion.  The only way to get anyone excited is to say that.  This may sound hard, but it is not.  There are many metrics that need to be satisfied for a battery, including, energy, power, charge time, cost, life, safety, low temperature and high temperature stability.  If you think that you have something that looks better in any one of these, you are doing better than Li-ion.  Cost is the easy one if all else fails.  You can always safely say something like “our preliminary cost estimates suggest that the battery will cost less than something-small/kWh”.  Other end of the spectrum is energy, which is the hardest.  If you go down this path you really need Step 6. 

Step 6:  Always confuse energy with power.  It’s completely appropriate to say “Our pixie dust battery can discharge a factor of 10 faster than Li-ion, therefore EVs based on pixie dust have a longer driving range comparable to Li-ion EVs”  or “our batteries can be charged in 5 minutes, providing more energy than any battery known to man or aliens. On a separate note, we only seem to get one cycle from our battery; we think this has something to do with aliens” 

Step 7:  The paper is ready and it is time to submit.  Never send the paper to a journal that specializes in publishing papers in batteries.  This will get your paper into the hands of traditional battery-types who remember past history, know what works and what does not, and have a strong scientific foundation in the field.  Such knowledge can be an impediment to your out-of-the-box thinking.  Remember Step 2.  Always choose a journal that is disconnected from the battery field. 

Step 8:  With the paper coming out, it is time to prepare for a press release.  Remember that the press wants to hear that this is a breakthrough.   So despite what the peer-reviewed paper proves, make sure you call it a breakthrough at the press release.  Remember that Steve Jobs did not really have a working iPhone when he announced it to the world, and declared that they would ship in 6 months.  If it is good enough for Steve, I’m sure it is good enough for you.  So don’t be shy in telling the press how great your battery will be.  Make sure that you give interviews to numerous publications.  Remember our definition of a breakthrough:  you need multiple publications to say it is one.  So target many outlets. 

Step 9:  The day has arrived; the publications are out; and you have spend the better part of the day googling yourself to see how far the word has spread.  This is the day when you may hear skepticism (some contained within the articles and others in emails addressed to you).  Remember Step 2.  Remind yourself that the iPhone had many critics (e.g, the proximity sensor will not work.  Who would want to surf the web on a phone anyway? Atleast they got the first problem right!).  If it worked out for Steve Jobs, then it could work out for you too.   

Step 10: Remember “Practice makes perfect”.  So go back to Step 1 and repeat. 

All the best. 


1.  Now, you may search Wikipedia, or read some articles that claim that the earth was known to be round since before the Common Era.  But that is because you are reading the literature.  Did we not drill into you in Step 1 that this was bad!   Now stop looking up stuff and get with the program.    

Tuesday, November 10, 2015

Developing a roadmap for energy storage deployment

In my post titled “A boom. Then a bust. And now, a new equilibrium?” I had argued that this was a very unique time in batteries.  My opinion is that there are two fundamental trends that we need to pay attention to:

1.   Batteries are getting increasingly deployed, both in vehicles and on the grid, despite high costs.  

2.  Costs of batteries are coming down.  Typically, costs fall at 5-6% per year.  With the present push by the big companies, costs are expected to come down by a factor of two within the next 5 years. 

Let us talk about the first trend:  We have always had batteries in phones and laptops, but in the last few years, they have come on their own in bigger applications.  And this is despite the higher costs. 

Nissan has decided to commercialize a cheaper car with less battery (and, consequently,  less range), while Tesla makes a big car with a big battery to kill range anxiety (but adding wallet anxiety?). 

Utilities all across this country are deciding to install storage on the grid to learn more about how they work and how to monetize them, while home-owners with a green thumb (and a wallet to match) are thinking about how to stick it to the big guy. 

All this means that we are learning, everyday, about how the batteries are working in the real world.  Do they cycle well?  How does the fade change with temperature?  Are we going to get our money back if we buy a battery versus building a transmission line?  And what is the actual cost of the Tesla Powerwall per kWh once you add the invertor and pay for the labor? 

This is great for everyone.  Deployment will be the key to figuring out what works and what does not.

And this trend will only increase with the second trend of decreasing costs.  Many more will decide to go for an EV and install batteries with solar panels.  We will plug the EVs on the grid to try to learn if they can help pay for the costs.  We will plug them in all at the same time and we will know how the grid will react. 

But reality is that batteries will still be expensive, maybe a factor of two more than where they need to be. They will still have less life than solar panels.  And less energy than gasoline.  And will be less safe than Niagara falls (well… I suppose that depends on how hard you want to kill yourself!).   We will need new systems to satisfy these gaps. 

But new systems will require time and money to reach the market.  Both of which become scarce commodities when big players are cutting the costs of Li-ion batteries dramatically. 

This was the context for the day when more than 200 of us met at the 2015 Bay Area Battery Summit on Nov 3 at Berkeley Lab.  It was a great day with insightful talks and panels, wonderful hallway conversations with thoughts leader from academia and National Labs, industry, and policy makers. 

The main theme was to explore the interplay between technological innovations in batteries; the changing market, and the role policy can play in accelerating deployment.

I thought I would use this blog as a way to list a few key discussion points that I heard.   These are fodder for future blog posts, so I will keep them short.

- Storage resembles the solar market in the late 2000’s.  But there are differences (and significant ones at that).  What can we learn from solar?

- What does the equivalent of the California Solar Initiative look like for storage?

-  If reality is that the battery has to do more than one thing for it to be cost effective on the grid, what are those low-hanging use cases? And how can we get policy to align with this reality?

- What is the balance between deploying what we already have versus finding new things that can solve the cost, life, energy and/or safety challenges?

- If we do find something new, the reality is that we need 10 years and $300M to get it to market.  And on Day 1, the price of that battery will be huge!  How do these technologies compete with an existing technology where depreciation is already in play?

- How do we solve the conundrum that end users want to see data on a real system, while most startups can only make a few small cells? 

- Should someone in the midst of starting a battery company attempt to stay within the Li-ion manufacturing framework (to ensure that the big players buy them out)?  How can they try to disrupt the existing players and avoid the risk not being to get to scale?

- What are the role of the Federal and State governments, and the role of Universities and National Labs in ensuring that we have a portfolio of technologies in the market?

- And, the impossible problem: How do you get battery companies (and researchers) to stop overhyping what they do?  Funny enough, I was writing a blog post about this last week, but got distracted by the event.  Will get back to that next week.

As you can see, a VERY busy day with very deep conversations.   Over the next 2 months we will be consolidating the answers to these questions and providing a roadmap for ensuring success in the battery space.

One of the speakers (who, in the interest of fairness, will not be named) felt that batteries were somewhere around the “peak of inflated expectations” in the hype cycle.  Resetting expectations and doubling down on innovation and deployment will be crucial in moving towards productivity. 

But first, we need to talk about (and address) the problem of hype in batteries.  This will be the subject of my next couple of blog posts.

Stay tuned.


Sunday, November 8, 2015

Batteries at the cusp of a revolution?

I'm almost done with my next blog. Will post it tomorrow.  In the meantime, I wrote an op-ed on Greenbiz titled "Is battery technology finally at the cusp of a revolution?".  See https://www.greenbiz.com/article/battery-technology-finally-cusp-revolution
Check it out.

I was plugging the Bay Area Battery Summit in the op-ed.   Now that was a fantastic event.  Stay tuned for my blog post on the meeting (tomorrow).

Got to see a LOT of my colleagues from the battery industry.  That was fun!

Here is a shot from the event with the BMW i8 that the good folks from BMW brought in for us to see.  You can also see a fuel cell Hyundai behind it.


Tuesday, October 27, 2015

The Hero with four faces: Part 2

In this post we continue looking at the best-known battery chemistry for each of four metrics of importance:  energy density, charge time, life, and cost, aka, a “Hero” battery.  In Part 1, we looked at energy density and charge time.  Today, we will look at cycle/calendar life and cost.

And oh… there really is another metric that is also important (arguably more important!): safety.  I talk about this metric at the end.    

The cycle/calendar-life Hero:  Many battery people will argue that the Edison battery (yes, that Edison) is the Hero for cycle and calendar (or shelf) life.  And rightfully so.  I have a different Hero in mind, but, I can see why others think of the Edison battery.  I have heard that there are 30-year old batteries that still work!  And they cycle thousands of times, even tens of thousands of times.    

Family heirloom, anyone?

The Edison battery has a nickel hydroxide positive electrode (or, as battery people have been calling of late, cathode) and an iron negative electrode.  The positive electrode undergoes intercalation of protons, which I like (If you want to know why, read “A Brief History of batteries-Part 1 and while you are at it, read Part 2 also); it undergoes no disruptive structural change, although there is lattice expansion that occurs due to the co-intercalation of the water and potassium ions.  There is a side reaction wherein oxygen is evolved throughout the charge/discharge cycle, but, for reasons that I will get into soon, this is not an issue.  This electrode is virtually indestructible. 

The iron electrode undergoes a reaction by which the metal anode reacts with the electrolyte to form an hydroxide layer.  Lithium batteries have a similar reaction, by the way.  The Li metal reacts with the solvent and forms a layer.   Except that the layer is not reversible.  Luckily, the iron hydroxide is very reversible.  So much so, that we can cycle it thousands of times. 

So why are we all not taking to our lawyers about changing our living will to bequeath a nickel-iron battery to our children?  For one, the energy density is kind of low.  Very low!  At 25 Wh/kg, capacitors start looking interesting!  

 There is a second problem.  The battery requires regular topping off with water.  Like.  Every. Week.

Similar to the old lead-acid batteries.  The positive electrode of the lead-acid battery evolves oxygen during operation (so does the nickel-metal hydride battery).  The oxygen comes from water electrolysis.  Any oxygen that is released is removed from the battery via the vents.  A few weeks of this and the water level decreases.  Hence we fill in deionized water. 

Obviously, everyone hated the tedium of opening the hood, opening the vent, checking the density, and adding water.  When I could be using those precious 3 minutes to watch Gotham instead.   The humanity!  (well.. ok, there are some corrosive chemicals to deal with, so don’t try this at home). 

In lead-acid (and in the nickel-metal hydride battery) we solved this issue by “sealing” the battery.  This forces the evolved oxygen to migrate to the negative electrode and recombine to form water again.  Ergo, oxygen comes out, but nothing is lost.  Except electrons.  The battery self-discharges, but no capacity fade is observed. 

Why does this not apply to the Ni-Fe battery?  Because the iron electrode also breaks down the water and evolves hydrogen.  Hydrogen, unfortunately, does not recombine (at least not in any appreciable amount).  So any hydrogen coming out is lost through a vent.  So the water will dry up and you will be missing 3 vital minutes of insert-favorite-activity-here. 

So what do you do with a low-energy density battery that requires constant babying?  You keep it in the basement and bequeath it to you children and hope they forget that you really have nothing of value left.

Or you use it to learn some lessons.  Like, intercalation is good.  Or side reactions (like oxygen evolution) are ok, if they can be reversed.  Or making layers (like hydroxide) is perfectly fine, if said layers are reversible.  Or, water-based systems are very forgiving, and they can be maintained very easily. 

Are the problems in the Ni-Fe un-solvable?  I actually think they are not.  Other systems, like the zinc electrode, have had similar gassing issues and folks found that mercury was effective in poisoning the hydrogen evolution reaction (in addition to poisoning a few other things).  

Or maybe we can automate the watering system (a battery-drip-irrigation system anyone?).

Despite this, I don’t consider Ni-Fe to be the Hero because this system is not that practical, atleast as we know it today. My Hero is a chemistry that is close enough to the Ni-Fe system.

It is the Ni-hydrogen battery. 

This is the battery in the Hubble Space Telescope.  Launched in 1990, the battery operated in space for 19 years after which it was changed during service mission 4.  These batteries cycle 20,000 times when cycled at 50-60% of its full capacity and have an energy density of 75 Wh/kg; close to other nickel-based batteries.  Now that sounds like a Hero to me. 

So why does it cycle so well?  The positive electrode is the same nickel electrode that I like so much.   The negative is an electrode that evolves hydrogen on charge and recombines to form water on discharge. 

Wait… did we not just conclude that making hydrogen is bad?  Like seven paragraphs ago.  Contradiction much?

In my defense, I only contradict myself between posts, not within posts. 

In the Ni-hydrogen cell someone really made lemonade out of that lemon.  They decide to use the gassing reaction as the reaction to store energy.  But then, the question is: How do we store the gas? 

Therein lies the genius of the design: The battery sits inside a pressure vessel.  When the battery charges, the pressure builds inside the pressure vessel, but it stays inside.  There are no side reactions.  When you discharge it reacts right back.  The nickel electrode still has the oxygen side reaction, but this stays within the pressure vessel and recombines.  Nothing is lost.  Like I said, genius.  Of all the innovations in batteries I have seen and read about, this is the one that I’m still impressed by. 

If only we could get our Li-ion batteries to cycle the way of these older systems.

All seven readers of this blog know that there must be a catch with this chemistry.  In this case, it is cost.  Adding the pressure vessel does not help.  I don’t have any hard numbers, but I’m told it is several thousands of dollars a kWh.

Which bring me to…

The Cost hero:  This is a hard one.  But let me start with a few irrefutable facts:
  • All battery companies will drive cost to less than $100/kWh sometime in the future, irrespective of the chemistry.  But, what if the battery is made purely of gold, you ask?  Trust me, it will still cost less than $100/kWh. 

  • This price will drop to whatever-the-number-needs-to-be, to be less than or equal to whatever-is-claimed-by-competing-companies.  

  •  The actual math behind the claims will be guarded in a fashion that would make the CIA proud and will consist of various confusing assumptions related to cost of materials, cost vs. price, claims about cycle life, assumptions about Indians dumping the gold hidden in their mattresses etc.  

Now that we have settled on that, let us talk about cost. 

One can define cost a few different ways: Either the capital cost of the battery or as a cost per cycle (or charge passed) over the life of the battery.  Cost drops with scale of manufacturing (think, Tesla’s Gigafactory) and as this scale increases the cost of materials starts to dominate the overall cost of the battery.  Packs cost more than the cell, but as the chemistry gets more complicated (read, Li-ion) the pack costs are significantly higher than cell costs compared to simpler chemistries (read, lead-acid).  And then there is cost and price.  How is one to define a Hero, far less identify one?

The cheapest battery one can buy is the lead-acid battery which costs anywhere from $80/kWh to $150/KWh (and more, depending on the complexity of the system).  Unfortunately, the cost and the cycling go hand-in-hand.  So deep-cycle batteries cost more than backup-power batteries. 

And how long would a deep-cycle lead-acid battery last?  The 8 years and one thousand cycles that some companies promise, or the 5 years and a few hundred cycles that most seem to get?  Too many variables to make a judgment.

Then there is the Li-ion battery where the cost is expected to drop and we are not sure where it will end up.   Three different cost estimates suggest that the costs will hover somewhere between $170-220/kWh at the system level.  So twice the cost of the lead-acid, but with twice the life?  Assuming that the predicted cost reduction pans out.  And will lead-acid also drop in the same time?  Probably.    

Considering where Li-ion battery costs are today (at $350/kWh or more), one could call the lead-acid battery the Hero. But let us wait for 5 more years and revisit this.  We are living in a very interesting time indeed! 

So let us forget about identifying Hero’s and let us get down to things we can learn. 

It helps to have cheap materials.  Things like zinc, carbon and manganese are good.  Lanthanum-nickel alloy and cobalt oxide, not so much. 

But we cannot get too carried away with this either.  One of the cheapest batteries, from a materials perspective, should be the sodium-sulfur chemistry (the cost of these two materials would in the cents/kWh range).  But to actually make the sodium and sulfur work in a battery setting requires solid conductors and high temperature operation.  The final cost: $400/kWh or more! 

It helps to have aqueous systems: the manufacturing is simple.  No need for dry rooms.  And there is a certain amount of system-level advantages to be gained when complex pack designs and battery management systems can be avoided.  So lead-acid-like design and manufacturing would help.

But cost is really cost per unit of energy ($/kWh).  So low energy systems don’t help.  Li-ion stands a gain because of this.  But the advantage of Li-ion (the high voltage compared to aqueous systems) is also the reason we need equipment with high cap. ex. 

If you did not know this already, there is no free lunch. 

These are the four Hero’s.  Each metric has a different Hero (lithium thionyl chloride, lithium titanate, nickel-hydrogen, and lead-acid).  High energy requires high voltage.  This tends to kill the charging capability.  If you want life, it is best to have reversible side reactions.  And cost… what can I say about that one! 

Hold on a second!  Clearly I’m not addressing the elephant in the room.   The one problem that I really ought to be providing insights about and identifying a Hero for: safety. 

A Safety Hero? Well… If you thought cost was hard, safety is even worse.  Batteries are energy storage devices where the oxidizing and reducing chemicals are stored right next to each other.  So asking me to identify a safety Hero is like asking which one of the unsavory characters in Pulp Fiction is the good one (The Wolf?  Haven’t you told your best buddies that you would help them move a dead body?)

I would have said that any system that does not have a flammable electrolyte would help.  But the lead-acid fire at the wind farm at Kahuku, HI would make that sentence a bit hollow.  Clearly, if you work hard at it, any battery can go up in flames!

Having said that, we have been using batteries day in and day out for a long time without anything bad happening, so one can learn some lessons from it.   

Clearly flammable electrolytes don’t help.  Materials that can result in exothermic reaction don’t help.  Bigger the amount of chemicals next to each other, bigger the risk.   So flow batteries will be inherently safer than contained batteries. 

If you ask me for a ranking, Li-ion would be the worst; aqueous chemistries are MUCH better; flow batteries would be the safest.

But safety is not just about fires.  Many flow devices have toxic and corrosive chemicals in them.  These devices have to be contained.

What this all means is that a lot depends on the battery company.  If care is taken to make the battery safe, they can be made safe.  But that costs money.  Companies that pay attention to this and spend the money to do it will find ways to keep the battery safe.

Summary:  So what do our four (or five) Heroes teach us?   That buying a battery is like buying a house; you have to compromise. 

Remember the Third law of batteries from my blog post The Three Laws of Batteries (and a Bonus Zeroth Law):  “Of the four metrics that batteries are graded on for a given application (i.e. performance, cost, life, and safety), typically, only two can be simultaneously achieved. If the battery is designed to also perform satisfactorily on a third metric, it will fail spectacularly on the fourth.”

That about sums up the dilemma.