Jeff Brown (aka WestTexas at TheOilDrum) and others developed a model they called the Export Land Model a couple of years ago to point out that the impacts of "peak oil exports" might be more abrupt than expected by most of the people who studied Peak Oil. The effect of domestic use by the oil producers could have a very significant impact on the availability of oil for export, and the timing of no exports from key producers could come sooner than expected.
The interest in ELM peaked and then dropped off as other events took the attention of many. Then Crash_Watcher published a series of three reports on ELM, starting with "An Export Land Model Analysis for the USA-Part 1" and continuing to Part 3. In the first Part, CW reviewed the basis and analysis for the ELM, especially as applied to the USA. In the second Part he discussed some of the more obvious scenarios that were possible, and predicted that some of the key exporters of oil to the USA could be out of product to ship within the next 20 years.
In the most recent Part 3 CW gets into the effect of EROIE (Energy Returned on Invested Energy) on this situation and concludes
Rather, the sum of exports from the top ten, as a group, are on a steady -6%/yr decline until hitting zero net exports in 2027. At this point, I assume that the USA is on its own and now is forced to live within it own means of producing petroleum. But by then, the ERoEI is headed toward 2:1 and by 2030, the USA’s predicted effective consumption is predicted to equal about 1 bbs/yr. That’s about half of the USA’s production rate of 2 bbs/yr—at a ratio of 2:1, the other half is expended on production. Thereafter, if the ERoEI stabilized at 2:1, and if the USA didn’t export any petroleum, then US consumption would also stabilize at about 1 bbs/yr.
Since today the USA is using about 6.3 bbs/yr, this is a sizeable drop in the next 16 years. CW says he will "give a summary of the results of this analysis and discuss what implications this has for the USA and globally" in Part 4, yet to be published.
I suggest you go read the first three Parts and comment to CW about what you see of his analysis. And be there to see his final conclusions.