These results are theoretical, and promising. The pics are stills from a upwelling calculator I built with xcode and cocoa on Apple OS 10.5 (Leopard) which is available on the reference page for download (Mac only), source code included. The calculator allows you to see the results ( upwelling rate, exit temperatures, heat transfer coefficients, Reynolds numbers, etc.) when the parameters are varied ( temperature profile, conduit sizes, conduit positions,wave height).
Pacific Results
Notice the hysteresis, which is a result of the heat conductance for turbulent flow (the up-flow) being more conductive than that for laminar flow. The white line is the performance as the wave height is decreasing, and the flow is tubulent. As the wave height increases from a calm ocean, there is not enough heat transfer to reach turbulent flow until about a wave height of 0.7 meters at which point the flow increases, becoming more turbulent and more heat conductive until an upwelling flow of 1.1 cubic meters per second is reached.
Atlantic Results
The Atlantic was not so easy as the Pacific, a result of the temperature profile having a higher slope. I found it necessary to move the top of the exchanger to the deeper level of 100 meters. When the exchanger had the same dimensions as the Pacific example there was not a transition from laminar to turbulent flow. Also, the data ends with a waveheight of 2 meters, because at that point the upwelling exit temperature is near that of the surface temperature, and the calculation becomes more difficult. One may assume, though that the upwelling rate cannot decrease as waveheight increases, and the trajectory of the graph will be similar to that in the Pacific.
Nutrients
In the north Pacific, the nitrate level at 300 meters depth has been measured at 17µM and the phosphate level at 1.1µM. One cubic meter of upwelled seawater from this depth will contain 1.2 grams of fertilizer consisting of nitrate and phosphate.