By Robert Campbell NEW YORK, Nov 21 (Reuters) - This week's stunning collapse in the price of West Texas Intermediate crude oil at Midland, Texas, is a stark warning to those betting on a narrowing Brent-WTI spread for the New Year that a lot can still go wrong on that trade. WTI at Midland sank $13 a barrel to change hands near $20 per barrel below the price of WTI at Cushing earlier this week as traders dumped barrels ahead of December pipeline scheduling. Sister grade West Texas Sour , which is still a component in some price formulas for foreign crude oil, suffered a similar meltdown. Dramatic moves indeed and worth closer examination. To be sure, the magnitude of the weakness seen this week is mainly due to the limited liquidity and volatility in U.S. physical crude oil markets ahead of pipeline scheduling. But the fact that such a huge decline can occur is instructive. While a few years ago it would have been remarkable for WTI at Midland to trade more than $1 below WTI at Cushing, discounts are getting larger and larger. What has happened? For a long time pipeline capacity out of Midland exceeded demand. So that was the case the price of WTI at Midland would be held by arbitrage forces at the Cushing price less the cost of shipping. But now pipeline capacity out of Midland is no longer sufficient to meet demand. Growing oil output in the Permian Basin has turned Midland into a "mini-Cushing." This is a critical point. When shipping capacity at a pipeline hub is in surplus, market power accrues to the oil producers who are able to charge top dollar for their product. Yet once pipeline capacity is no longer enough to clear demand market power swings violently to those with control over the capacity. Anyone with more oil than pipeline space is at the mercy of market forces. FROM MINI-CUSHING TO MEGA-CUSHING OK, no problem though, many may say. The problems at Midland this week are of a one-off nature and besides, help in the form of new pipelines that will send crude to Houston instead of Midland and Cushing is on the way in early 2013. There is nothing wrong here. The three-day period each month when physical traders balance their books and dump odd lots of oil always produce wild price action but very little oil actually changes hands. Moreover the extension of maintenance at Phillips 66's 146,000 barrels per day Borger, Texas refinery cut regional crude demand more than expected in November, leading to a surplus of oil on hand. And yes, there will be new pipelines. Lines out of Midland will add a combined 80,000 bpd in capacity to Houston and another 90,000 bpd to the Port Arthur area next year. Similarly, the Longhorn project near El Paso will allow up to 225,000 bpd of Permian oil to bypass Midland and Cushing altogether and move directly to Houston starting in the first quarter. And finally, the 250,000 bpd Seaway pipeline expansion will increase offtake from Cushing that should improve pricing at that hub. But hold on. All of this assumes that these new pipelines clear the market and return pricing power to the hands of producers. Notice that these pipelines all add a huge amount of capacity into Houston, most of which will be light oil. And this all comes as Eagle Ford crude capacity into Houston keeps rising. In other words, is there not considerable risk that the "Cushing problem" spreads from Cushing and Midland to engulf Houston as well? If that happens, will we not simply see light crude prices plunge in Houston to levels well below the global market? Indeed, Eagle Ford barrels and initial Seaway flows have already displaced almost all of the imported light sweet crude from Houston-area refineries, according to U.S. government data. Some 150,000 bpd of foreign light-sweet crude was processed in Houston-area refineries in August, far less than the expected flood of new barrels into Houston from Cushing, the Permian Basin and the Eagle Ford crude oil play. That suggests light sweet crude in Houston will have to compete on price to push out other foreign imported crude. How much of a discount is anyone's guess but due to restrictions on U.S. crude oil exports, local refineries will hold considerable sway over pricing. That in turn means big discounts could be needed to sway buyers, especially during periods of refinery maintenance. Next year's pipeline expansions will ensure a bigger market for North American light sweet crude, but will it be big enough? And if Houston is not big enough, can the market be sure that adding more Gulf Coast outlets will be enough?