🚨Preprint Alert🚨IT'S HERE!

For the first time, we use a large-scale controlled study of O&G #methane emissions to answer a basic question: How effective are commonly used Leak Detection & Repair (LDAR) programs?

There's a LOT we learnt here. Thread: 1/ eartharxiv.org/repository/vie…
EPA has proposed regulations to reduce #methane emissions from O&G sector. If global energy sector methane were a country, it would be the third largest emitter behind only China & US.

Key part of these regs are leak detection & repair (LDAR) programs. 2/ washingtonpost.com/climate-enviro…
LDAR program are where O&G operators survey their facilities with IR cameras to find and fix leaks (see NYT article for videos).

But, we didn't really know if these LDAR programs are effective in reducing #methane, beyond anecdotal evidence. Until now! 3/ nytimes.com/interactive/20…
Why? Because it's hard. Need to meet 5 conditions:

1. Control group where leaks are not repaired
2. Treatment group where leaks are repaired
3. Measure all emissions incl. vents, not just leaks
4. Repeat measurements
5. Avoid selection bias (select only low emitting sites) 4/
But that's what we did.

We randomly selected ~180 O&G sites in Alberta.

We split them into 4 groups - 1 control (C) & 3 treatment (T1-3) - with ~45 sites each.

Control - operators not informed about leaks.
Treatment - operators asked to fix leaks.

Survey schedule below. 5/
To track emissions over time & across surveys, we put physical tags on leaks. When operators fix leaks, they note a 'date of repair' on tag (that's how we know leaks were fixed).

No tags were left on leaks at control sites.

Based on these tags, we have 4 treatment effects. 6/
We find repairs are highly effective. Look at comparison between tagged leaks that were not repaired vs. tagged leaks that were repaired.

Gray: initial survey
Pink: follow-up survey

1) After repairs, emissions ~0.
2) Even if not repaired, emissions do not grow with time. 7/
At treatment sites that were repaired consistently, average number of leaks reduced by 50% (5 to 2.5).

At control, little change (2.2 to 2.0) - slight decrease possibly from voluntary maintenance.

At treatment sites where repairs were *not* done, leaks went up (1.3 to 1.6). 8/
What does this mean?
First, high survey frequency + repairs ➡️fewer leaks.

Second, voluntary inspection many reduce emissions, but reductions much smaller compared to LDAR programs➡️importance of regs.

Third, consistent repairs after surveys key to reducing emissions. 9/
Not all operators are created equal. You can see ~2 orders of magnitude difference in average site emissions across 18 operators. We can speculate why:

1) Different asset portfolio (more oil production, more emissions)
2) Different maintenance practices
3) Super-emitters 10/
Not all sites are created equal.

Oil sites emit, on average, 2x that of gas sites.
Sites with multiple wells emit, on average, 2x that of sites with single wells. 11/
Venting is a much bigger problem than leaks. Part of the reason is the way we classify many emissions as vents even though they can be fixed (e.g., open thief hatch on a tank).

This is why LDAR surveys can help find anomalous vents even if they are only targeted at leaks. 12/
Last thing: Tanks and pneumatics are the biggest sources of emissions. Look at distributions of tank and non-tank emissions and how they changed over the course of the year!

In 2019,
Avg. tank emission = 105 kg/d
Avg. non-tank emission = 14 kg/d, order of magnitude smaller. 13/
There's a LOT more in the paper! I was a @Stanford post-doc when this start, now I'm faculty at @UT_PGE! And big congrats to @Lyra_Wang, who led the analysis!

These are preliminary results. We are releasing all data so others can conduct more analyzes! /End
For those who asked: @drvolts @StevenMufson @dino_grandoni @ProfVonfischer @MikeLeeFW @johnupton @lorakolodny

This is a preprint and will undergo external peer review, but we wanted to get data & results out so insights can be helpful for #methane regs.

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Arvind Ravikumar

Arvind Ravikumar Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @arvindpawan1

26 Oct
I'm out here in WY near the Wind River Range to measure #methane emissions. As I write this, EPA is working on releasing updated methane regulations.

Let me tell you why this is not necessarily bad news. A thread on the complexity of #methane accounting: 1/
First, why is #methane fee a great idea? Recent data show that #methane emissions from operators vary by many orders of magnitude. So it makes sense that under a fee, responsible operators will be rewarded and those with high emissions will be penalized.
So, what's the issue with BBB proposal? Fee is entirely based on self-reporting to EPA GHGRP. 2 problems:

1) Emissions are underestimated in EPA report
2) Only facilities >25000 CO2e required to report

This sets up potential for fraud as it's easy to reduce emissions on paper.
Read 9 tweets
20 Feb
Good summary of what went wrong with the TX grid. I am going to try to explain what happened on the natural gas supply chain.

TL;DR: Combination of extended cold, unique basin properties, old pipes, and gas/electricity dependence. Thread. 1/
First, here's the natural gas supply chain. The parts that failed were in 3 areas:

1) oil & gas wells
2) Gathering lines
3) Equipment malfunction at power plants
4) Outage cut power to compressor stations that moved gas

+Other long-term issues like limited storage in TX. 2/
1) Why did O&G wells fail?

Permian basin is a liquids-rich basin. In addition to gas, wells also produce oil & water. For e.g., for every barrel of oil produced, you bring up 2-3 barrels of water.

In extended cold, water freezes and blocks the flow of gas from the well.
Read 9 tweets
11 Dec 20
Today's @Ben_Geman generate shows that #methane venting and flaring in TX and ND reached record highs in 2019.

This has serious implications for the lifecycle GHG emissions of natural gas power plants. US average leakage rate masks a lot of variation. 1/
So, Alan Strayer - UG student in our lab, painstakingly traced gas flow from production basins to power plants to estimate state-specific leakage rates.

US avg. leak rate is ~2%, but states in the Midwest/SW have far higher leakage while states in NE/SE are lower. 2/ Image
Part of this is because of high venting/flaring as @Ben_Geman reports, in the Permian & Bakken basins, but also Canadian imports.

Looking at power plants by state, we see that states in MW/SW have higher emissions intensity than plants in the NE. 3/ Image
Read 6 tweets
13 Oct 20
Folks citing direct employment numbers in the gas industry in PA to suggest Biden is wrong on his "no fracking ban" policy don't get the ecosystem of people in these communities. It's not just a job - it's family, it's small businesses, & entire communities that will be affected.
I have been in these towns, I collect data in these towns, & it's not hard to see how entire community revolves around an industry.

It's not just a job, it's a way of life.

To suggest that it's only 10K jobs or votes is patronizing, elitist, reductive, and importantly, wrong.
I am one of the few in this country to engage with this issue intellectually while also forging a personal connection with these towns & workers.

And I have come to understanding their perspective, their hopes & challenges, and how we should really talk about energy transition.
Read 6 tweets
14 Jan 20
🚨New Paper Alert🚨In a first study of this kind, we *empirically* show that leak detection and repair programs - a common methane policy tool - are indeed effective at reducing emissions over many years of implementation. We also found a few surprises. 1/ iopscience.iop.org/article/10.108… Image
First, emissions reduced by 44% between two LDAR surveys conducted over a period of 0.5 - 2 years from the initial survey.

Compare this to EPA (or other state policy) assumptions that annual surveys reduce leaks by 40%. Pretty close. But, note I said emissions, not leaks. 2/ Image
Important context:

Methane emissions consists of leaks (unintentional, fixable) and vents (intentional, not fixable). Leak detection policies only target leaks.

But, some vents can be fixed because they vent far more than what they were designed for. These anomalous vents. 3/
Read 8 tweets
7 Nov 19
I keep getting hate mails, so let me explain.

In any context, using "population control" or "population reduction" is never okay.

Your intentions may not be malicious, but given the genocidal history of that phrase, it is best to not use that framing.

Population growth in developed countries is below what's need to maintain a steady population.

So, when you're referring to "population reduction", you're specifically referring to the developing world.

Here's global fertility rate, where <2.3 is below replacement rate.
And saying population should be reduced when referring to the developing world has a long history in racism, eugenics, forced sterilizations, and other unspeakable horrors in our history.

This isn't new. There's a long history of well-funded Malthusian overpopulation alarmism.
Read 6 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!


0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy


3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!