I would like to try to add some signal to the noise in this thread due to, I think, some mis-communication.
One of the neat things about a siphon is that, although it's true that the ultimate output MUST BE LOWER than the source (that's the part about gravity which some of the posters were referring to), you can actually have the tube/pipe that the fluid follows temporarily rise ABOVE the level of the source - as long as the ultimate destination is at a lower point.
This will all only work if the source and destination are at the same pressure (or the destination is at a lower pressure than the source). If your destination has higher pressure than the source, the pressure may be high enough to overcome the force of gravity, and stop the siphon from working.
All that said, if the path which the fluid follows rises above the level of the source, you will need an active 'pump' to get the siphon *started*. Once started, it should continue to flow without additional pumping.
The reason this can work (the path going higher than the source) is that the total weight of the fluid on the 'low' side of the path, which leads to the output, will always end up being greater than the weight of the fluid on the on the 'uphill' side of the path (assuming that the pipe/tube is of constant diameter; if your tube/pipe narrows as it goes to the destination, this could also fail because there would be a greater volume/weight of water on the 'uphill' side).
As the weight of the water on the 'downhill' side of the path is greater, it will create a negative pressure difference on the fluid 'behind it' in the path, and assuming there is any kind of pressure at the source (e.g. atmospheric pressure), the atmospheric pressure + the pressure difference caused by the weight of the water on the 'downhill' side of the path, will force the fluid in the uphill portion of the path to flow uphill.
So, what all this means, is that if you have a pipe running from the ocean, to an 'entry' point +9m above sea-level, but then the pipe continue down to below sea-level (so that the ultimate destination point is below sea-level), then YES, you should, maybe, be able to get a siphon going, and not need to actively pump the water.
HOWEVER, there's another wrench in the works - remember I talked about pressure earlier? When water interacts with magma, it will turn into high-pressure steam. It's quite possible that your siphon WON'T work (or at least won't move fluid down at the rate you'd like), because you have high pressure steam at the bottom of that pipe, which could force it's way up the pipe until there is an equilibrium between the steam pressure and the weight of the water (as the steam expands up the pipe, the pressure will drop; depending upon the initial pressure of the steam, it might reach an equilibrium, or it might race up your pipe all the way to the ocean and then start billowing steam out the 'inlet' of the pipe).
I think the pressure issue from the steam, may be the reason that some geothermal plants actively PUMP working fluid down underground - because they need the pump pressure to counteract the steam pressure.
