Water Usage in an Oil Refinery

There has been much controversy regarding the amount of water used to produce a gallon of ethanol. Considering just the usage in the ethanol plant (ignoring any irrigation requirements for the corn), this amounts to about 4 gallons of water per gallon of ethanol produced:

New Research Paper Finds Water Availability Critical to Growth of Ethanol Industry

Generally, an ethanol plant will use 10 gallons of water per minute for each 1 million gallons of ethanol produced. A typical 50 million gallon plant, would need 500 gallons per minute of water.

There are no publicly available records on water use by ethanol plants in the United States, the authors found, with the exception of the Minnesota Department of Natural Resources.

Minnesota plants use a range of 3.5 to 6 gallons of water per gallon of ethanol. Average water use has declined from 5.8:1 in 1998 to 4.2:1 in 2005.

Authors of the paper said 4 gallons of water per gallon of ethanol is a good estimate with the current technology.

I have frequently been asked how this compares to the water usage for an oil refinery, and each time I do some back of the envelope calculations and come up with about 0.5 gallons of water per gallon of crude oil processed. But the question comes up often enough that it is worth documenting.

According to an article in the February 18, 2007 Billings Gazette:

Here are the top users of the Billings Public Utilities Department

TOP WATER USERS (GALLONS PER YEAR)

1. Billings Heights Water District, 848 million.

2. ConocoPhillips Refinery, 456 million.

3. PPL Montana, 53.4 million.

4. Casa Village Mobile Home Court, 49.6 million.

5. St. Vincent Healthcare, 39.6 million.

The ConocoPhillips refinery in Billings processes 62,000 bbls of crude oil, or 2.6 million gallons per day. The reliability of most refineries is in the 90-95% range, so if we assumed 92.5% on-stream time, the refinery processes 2.6 million * 365 * 0.925, or 879 million gallons of crude oil per year. The water usage then amounts to 456 million/879 million, or 0.52 gallons of water per gallon of crude oil processed.

Note that this is actual make-up water that is brought into the refinery. In other words, this is the actual usage of the refinery.

23 thoughts on “Water Usage in an Oil Refinery”

  1. Eight times the water. Oof!

    Thanks for penciling this out. This is a particularly good statistic to trot out, when discussing the environmental externalities of ethanol. We can talk about soil loss all day long, but that’s harder to quantify and less meaningful to the average person. Water usage, at least, the average joe can understand.

  2. I had calculated the number several times for people privately, but I just saw that story listing the water usage of the Billings Refinery today. I realized I could document an actual public source that can be referenced, instead of always giving people a private communication.

    What most people probably don’t appreciate is the rate at which fossil aquifers are being pulled down. It’s one thing to use a lot of water, but it’s quite another to use a lot of fossil water.

    Cheers, Robert

  3. How recoverable/reusable is the water used in EtOH fermentation and petroleum processing? Related to this, does a refinery discharge any (treated?) water into the waste stream?

  4. How recoverable/reusable is the water used in EtOH fermentation and petroleum processing? Related to this, does a refinery discharge any (treated?) water into the waste stream?

    For both a refinery and an ethanol plant, the waste water is treated and discharged (although a refinery also has pretty decent losses from evaporation in cooling water towers). The usage numbers mentioned are makeup water that has to be replenished into each plant.

  5. Following on to Brian’s question, do you know enough about the ethanol process to make an educated guess about the potential to reduce the required makeup water by re-using treated process, recondensing water removed in the distillation process, or similar measures?

    My assumption is that, thus far, ethanol production is sufficiently small scale and sufficiently immature that they don’t use enough water, in an absolute sense, to be forced to conserve, whereas the refineries reached that size scale and maturity a long time ago.

  6. My assumption is that, thus far, ethanol production is sufficiently small scale and sufficiently immature that they don’t use enough water, in an absolute sense, to be forced to conserve, whereas the refineries reached that size scale and maturity a long time ago.

    I don’t think that’s it. Oil and water don’t mix, so it is quite easy to separate out water in an oil refinery. In an ethanol plant, the products are all highly soluble with water, making it more difficult and energy intensive to achieve a high purity stream for recycle.

    But, I think the biggest source of lost water in an ethanol plant is that the mash left over from the fermentation is saturated with water. That mash of often fed wet to cattle, so there really isn’t much hope of getting that water back.

    RR

  7. Not arguing with your calculation, but just think that if you quote the water usage for ethanol based on gallons of ethanol produced, that the oil refinery water usage should be provided in terms of gallons of gasoline produced (not gallon of crude oil processed, or else you would need to base ethanol on amount corn [or other crop] processed, correct?). So you can compare on a similar basis.

    If you did that and assumed that 19.5 gallons gas produced per barrel of oil (figure I see cited several places on other sites anyway), then the figure would be ~1.12 gallons water used per gallon gasoline produced, right? So ~4X difference vs. ethanol.

  8. If you did that and assumed that 19.5 gallons gas produced per barrel of oil (figure I see cited several places on other sites anyway), then the figure would be ~1.12 gallons water used per gallon gasoline produced, right? So ~4X difference vs. ethanol.

    No, because you get more than gasoline out of that barrel of oil. You also get diesel, jet fuel, sometimes propane, butane, and other compounds (depending on the processing). That’s why I did it on the basis of a barrel of oil. About 95% of the barrel ends up as fuels. And of course if you do it on the basis of gallons of water per BTU of fuel produced, it gets MUCH worse for ethanol.

    RR

  9. Understand your comment about producing fuel other than gasoline, but in order to compare the ethanol water usage you provide to the oil refinery water use, it would still need to be based on the amount of fuel produced (regardless of whether that also includes diesel, etc.) instead of on raw material consumed. Any analysis needs to compare on a similar basis (i.e. fuel produced) or the comparison is not meaningful.

  10. Any analysis needs to compare on a similar basis (i.e. fuel produced) or the comparison is not meaningful.

    That’s why I indicated that 95% of the barrel of oil ends up as various fuels. If you want to make the comparison on a product output basis in both cases, then divide the 0.52 by 0.95. You get 0.547, not that much different. That’s why I didn’t worry about it, and why I did the simpler calculation. I knew it wouldn’t make much difference.

    Cheers, RR

  11. I am interested in locating any published data that might be available on water used in petroleum refining in regards to this very issue – water needed to produce gasoline vs. water used to produce ethanol. Do you know where I might be able to find that? Not that I don’t trust your numbers – I just need a source I could use to verify them.

  12. I am interested in locating any published data that might be available on water used in petroleum refining in regards to this very issue – water needed to produce gasoline vs. water used to produce ethanol.

    Well, that was published data. The water usage number and the refinery capacity numbers are both in the public domain.

    Cheers, Robert

  13. Great discussion. I have been wondering about this myself. Siemens (who make the water purification system for ethanol plants) tell me the water usage ratio has been brought down to 3:1 from 4:1 in the last year. But they dont see it going much lower than that due to cost.
    Ethanol Producer magazine had a story the other day saying that water usage could be brought down to 1.5:1 or even zero, using recycling and recovery technology, but it sounded like it be be costly.
    David Adams
    The Fueling Station blog

  14. How does DDGS fit into the water use calculation? In the petroleum example water is allocated across the different fuel products, but in the ethanol example all of the water use is billed to ethanol.

    On a simple mass balance, about 1/3rd of the water use could be billed to DDGS. If a displacement method is used to allocate water use, I suppose you’d need to figure out how much water is used to grow the corn (and other ingredients) displaced in animal diets by the DDGS.

    Thanks,

    Mike

  15. Much thanks for the post. Someone was claiming that it took over 1000 gallons of water to refine one barrel of oil, and when trying to see if that was real or not, I found this entry of your blog.

    I was a little disappointed that your reference for the capacity for the Billings refinery was a wikipedia entry that was flagged “needs citation”, so I found
    http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/refinery_capacity_data/current/table5.pdf
    which says the Billings plant refines 58,000 barrels per calendar day. That doesn’t change your numbers much. (I suppose there are fewer working days than calendar days per year and perhaps the 62,000 bbls/day is for working days?)

  16. Lets try and get all the proper information out to the general public and not a bunch of bits and pieces which tilt the argument one way or another. Speaking from the ethanol stand point the process requires about 0.5 gal water per gallon of ethanol produced. The rest of the water usage in an ethanol plant come from cooling tower evaporation and blowdown as well as boiler blowdown. I know these are factual statements since I have been designing ethanol plants for the past 13 years.

    Now, you stated that the refineries use about 0.5 gal water/gal of processed crude oil, however these numbers do ot include the cooling towers or the boiler blowdown. Lets get these numbers into the equation and then make our comparison.

    Tony – Ethanol Designer

  17. I’ve been looking for a good number for quite a while now for water consumed to create a gallon of gasoine. You can find numbers all over the map, all over the web.

    I keep hearing a number quoted by a Conoco-Phillips rep verbally of 8 gallons of water consumed per gallon of gasoline.

    I looked up the BP Whiting refinery.
    Daily barrels of crude: 400,000
    Daily water use: 140 million gallons
    http://whiting.bp.com/go/doc/1550/165406/

    At 44 gallons of refined product per barrel of crude, that works out to 7.95 gallons of water per gallon of gasoline. So… maybe that 8 figure is correct.

    Marc
    http://www.itsgood4.us

  18. Now, you stated that the refineries use about 0.5 gal water/gal of processed crude oil, however these numbers do ot include the cooling towers or the boiler blowdown. Lets get these numbers into the equation and then make our comparison.

    Of course they do, since that was the water usage by the refinery. Water lost to blowdown has to be replaced. The usage numbers are correct.

  19. I was a little disappointed that your reference for the capacity for the Billings refinery was a wikipedia entry that was flagged “needs citation”…

    I was actually working there at the time, so I knew the number I quoted was correct. You could probably find it in a number of other locations. But the refinery was beyond the 58,000 bbls/day level that you quoted at the time of that water usage.

    RR

  20. I keep hearing a number quoted by a Conoco-Phillips rep verbally of 8 gallons of water consumed per gallon of gasoline.

    Again, I have quoted actual usage numbers in the public domain.

    I looked up your BP link, and that isn’t a valid comparison of water usage. That is just water that they run through the plant as a once through cooling medium and discharge it right back into the lack. It isn’t actually “used” in the way the water in the Billings refinery (subject of this essay) or an ethanol plant is used since it never makes any contact with product. In other words, it is never contaminated so it never requires treatment.

  21. One gallon of ethanol on average requires five gallons of water. Not only that, ethanol is an expensive fuel to produce. Currently the U.S. government subsidizes its production at .54 a gallon. That’s to drop to .47, due to the 2008 Farm Bill.
    Most plants today are 50 million gallons. At five gallons per, that’s 2.5 billion gallons of water for one year of ethanol production. That water typically comes from regional aquifers that are quickly drained. While that frequently hurts regional wells, particularly rural residential wells, that’s not going to stop these plants.

    Energy production is a high priority in our nation. So that even in the Phoenix desert, companies can justify building an ethanol plant.

    Great post!
    CHeers!
    Georgina

  22. What about the amount of water required to actually get the oil out of the ground? That needs to be addressed in your totals, otherwise you are comparing apples to oranges.

  23. Chris,

    Generally speaking, the water used to get oil out of the ground is non-potable. Most of the time seawater or produced water is used. In the case of corn irrigation and ethanol production, it is fresh water. So the comparison in this post is apples to apples: Fresh water usage.

    RR

Comments are closed.