Discover more from ESG Hound
The Power of the Permit - Due Diligence on Steroids
...plus Notes on ENVX
I’ve written extensively in the past on my research using permitting documents. I’ve been guarded on talking about this topic because it’s one thing I know how to do that no one else in the investment research community understands. I'm a neophyte When it comes to identifying fraud on a balance sheet. Learning, but still not there yet.
I’d point you towards the power of this skill by looking at my work on SpaceX’s proposed (and subsequently pulled) proposal to build a gas processing plant, LNG unit, and power plant in Texas. Likewise, my research on Sustainable Fuel Greenwashing Scheme Gevo gets into the weeds on permitting as a tool.
Any research that can get you an edge on a company’s operations is obviously powerful. We know large funds pay lots of money to track oil operations with barge trackers and drones to see how full petroleum storage tanks are at major terminals. Funds have been paying people to monitor Tesla factory output since 2015 by counting cars leaving their facilities. It makes sense because a couple of thousand cars below or above expectations can cause wild swings in equity prices, especially on richly valued battleground stocks that trade on hype and expectations.
Whither The Permit?
I’ll start this post out with a caveat. Most of the guidance in this particular post is very specific to US manufacturing. In truth, there are tools to do similar work abroad. Several years ago, I completed a project for a client analyzing aluminum production in Europe that relied heavily on permitting data and emissions inventories on scrubbers used at smelters and foundries. But regulatory schemes abroad tend to be much more opaque than those we see in the States.
Regarding what kind of permitting can be used, there’s an entire buffet: Air, Waste, HAZMAT, and Wastewater are the big ones. We’ll get into more specifics in later posts, but today I want to focus on the basics of using Air Permitting in the United States.
Why Air Permitting is Such A Goldmine
Most air permitting requirements are driven by the Federal Clean Air Act
Because the Clean Air Act is the primary driver of most permitting requirements, you can analyze permitting documents in most of the United States using a similar framework (with wrinkles). Permitting, especially on so-called “Major Sources” is standardized to a very high degree.
Permitting itself is completed by local and state agencies
Completing FOIA requests (from Federal agencies) is a complete nightmare. Most Clean Air Act Permitting is done by state and local agencies under delegated Authority from EPA. State Open Records acts are generally much quicker and less contentious than FOIA. Many states even have portals you can pull up permitting documents without making a request at all (THANK YOU, STATE OF TEXAS)
Air Permits use standardized equations and methodologies to be authorized
We’ll go into this one a bit further down the road, but because there is an engineering and regulatory “Best Practices” guide to calculating air emissions, we can apply our knowledge of the technical process to find really interesting stuff in permits.
Air Permitting Docs are largely required to be in the public domain
Public records are great! The requirements for permits, applications, and engineering evaluations to be in the public domain mean we have lots of data to work with. A treasure trove of data with which to do completely legal corporate espionage.
Facilities have to permit at Design Capacity
This is my favorite part of Air Permitting. If you want to build a factory that is capable of outputting X amount of power, oil, or product, you have to base the permit application on the DESIGN capacity, not just how much you want to run it in order to get below emissions thresholds. I can’t tell you the number of times I’ve found a permit that disagrees with a fraudy company saying “WE’RE RAMPING MORE PRODUCTION THIS QUARTER!”
What an Air Permit Can Tell Us
Let’s take a look at a very simple block diagram for a widget factory:
Material goes in, and a finished good comes out. Pretty simple, right? The design of the factory, logistics, and operations will affect the number of widgets that can be made. The factory efficiency (minimizing spillage, for example) will determine how many widgets can be made per unit of input material. Let’s say, as designed, our factory can make 2.8 widgets per pound of steel that comes into the factory, an improvement over our competitors’ 2.5 widgets per pound.
This is the start of a mass balance exercise. Let’s fill that in:
Now, since no factory is 100% efficient, let’s say that each widget weighs 5 ounces. 28 million widgets times 5 ounces is 8.75 million pounds. We’ve got 1.25 million pounds of steel that came into the factory but didn’t come out as a finished good. It’s easy to call this waste, but our simple factory only uses raw steel, so the little scraps that fall onto the floor or are dropped in a bin after milling is just a by-product that can be sold back to a steel mill. That total amount of steel that we can recover for sale adds up to 1 million pounds:
But hang on!
8.75 million pounds (goods) + 1 million pounds (scrap) = 9.75 million pounds
Where did that missing 250 thousand pounds go?
Whoops, I forgot to mention that the milling process uses laser cutting and grinding that creates a cloud of fine dust, meaning that 250,000 pounds of dust is lost as an air emission. Well, sort of. The 250,000 pounds go into a dust collector/baghouse with 95% capture and 99.5% control efficiency. The mass and materials balance looks like this:
Because the cutting and grinding happen on the factory floor and a big vacuum diverts the dust, not all of those “fugitive” emissions make it to our control device (the Baghouse). 5% of the emissions go straight to the atmosphere, and 95% go to the baghouse (collection efficiency). The baghouse (basically a bunch of special sock-like filters with a blower) collects the dust. The bags/socks, as they fill up with dust, are shaken off by the machine and collected in a bin for disposal. But no control device is 100% efficient because of passthrough/leakage so we assume that 0.5% of the dust that goes into the baghouse escapes into the atmosphere.
All told, 94.53% (95% * 99.5%) of the emissions are controlled, and we have actual emissions of 13,687 lbs of Particulate Matter (PM10 or PM2.5) emissions to permit with our local permitting authority.
OK! Let’s put it all together:
This is the kind of information that permitting agencies get from companies seeking permit approval. But, those dastardly companies don’t want everyone to know everything about their factory so they will use Confidential Information protection to prevent the special manufacturing secrets from getting out. So they have their lawyers do this to documents released to the public:
Curses! We don’t know anything now except for the pounds of emissions at the end of the stack, disclosure of which is required by law.
Except that’s not true. We know a lot. Because air permits are based on EPA and local regulations, we know the required control efficiencies. We know the calculations used to estimate the amount of dust generated based on inputs. The specific nature of the emissions standards are written into rules, AND the means to calculate the pre-control device emissions come from standard sources like EPA’s AP-42 engineering manual.
We can reverse engineer factory inputs, outputs, and spillage to a high degree of certainty using just a handful of numbers. If a solution output is given and we know the mechanisms, regulations and formulas from which that output is generated, we can go to town!
Usage in the real world
Last week, I published a piece on battery stock promotion Envonix. In that case, the entire permitting document was unredacted. I was able to simply pull the “units per hour” charts from their permit application:
A more savvy company, like Enovix’s Fremont neighbor Tesla, would redact the heck out of their permit application. Tesla loves redacting permit documents (good for them, seriously). Their 2020 permit application for GigaTexas had some of the most absurd redactions I’ve ever seen:
Now, redacting the firing rate for a natural gas-powered burner is a new one for me in all of my years doing this sort of research. Especially because PM and SO2 emissions for natural gas combustion are a simple multiplication of the firing rate and a fixed value in AP-42. But long story short, the initial production rate for the factory (redacted by Tesla) was 350,000 model Y cars and 150,000 Cybertrucks, which I determined fairly easily using EPA requirements for car manufacturing and other process data provided by the company.
Back to our friends at Evonix: as I noted last week, the lines at their Fremont facility are the smallest variety of batteries supposedly being produced. But even if we didn’t know this, I could use the emissions rate from the electrolyte solution to figure out how many MWh of batteries they could make with the equipment. Here’s how:
If we know the Emission Rate for volatile organic carbons (VOC or POC) and the capture and control efficiency (we do, they have to be disclosed by law) then ==>
We know how many tons of VOC are being applied.
If we know this and what the VOC % is in the electrolyte solution (we do) then ==>
We know how much electrolyte solution is being used per year
If we know how much electrolyte solution is being used, we can reference patent documents and our knowledge of basic materials science/battery chemistry to ==>
Know How Many MWhr of battery cells they intended to make
All of this from emissions data, technical know-how, and an intimate understanding of how Permitting works! Wow
This won’t be my last post on Air permitting, but I hope you enjoyed it. Until next time…