It content so many problem to generate electricity from it...i.e temp,pressure,civil design so as my point of view we can but it will not feasible.

Generating electricity is important when it is economical !!!!!!!!!.Your question is like,"can we generate electricity from the water sent to cooling tower ???????".

Rest assured, this problem (scavenging waste heat) has not been overlooked by the engineering community. For example, gas turbine electrical power plants often recover much of the exhaust heat for steam and hot water production (ie; co-generation). If you have an automobile with a turbocharger installed, the turbocharger is scavenging waste heat from the exhaust to compress the air going into the intake manifold. There are many other examples.

Basically, whenever there's a temperature difference (no matter how small) one can, in principal, extract energy. One thermodynamic principal is that the maximum theoretical efficiency is proportional to the temperature difference (Carnot's principal, or the Second law of Thermodynamics). So, if the temperature difference is small, so is thermodynamic efficiency. Economics also play a role here. If a process has low efficiency, it's also likely to be un-economical.

I don't know of any examples of direct electrical generation from waste heat from an industrial chimney. I'm not saying there haven't been some installations out there that do this, but given the low temperature differences and small size, I suspect this is not a very efficient way to recover waste energy.

There are some solar energy generation designs based on this 'chimney' effect. But, because they all exploit relatively small temperature differences, they're all relatively low efficiency. Thus, the economics of the design is rather tricky. It's only through the economies of scale and free solar energy that the low thermodynamic efficiency can be overcome. Try searching Wikipedia for "Solar updraft tower".

As far as I know there have been attempts with using thermoelectric generators.

http://dailyfusion.net/2013/07/efficient-thermoelectric-cell-for-waste-heat-recovery-developed-14890/

http://www.komatsu.com/CompanyInfo/profile/report/pdf/164-E-05.pdf

http://www.gizmag.com/printed-thermoelectric-generators/26918/

You can find several other by Google using the keywords waste heat thermoelectric

Joachim's example brings up the topic of how ignorance of the second law of thermodynamics can lead to logical impossibilities. Not in Joachim's case, rather the folks at Wow Energy. In America, we call that 'snake oil.' Despite Wow Energy's assertion of near 100% recovery (efficiency), I'd be shocked to see their design exceeds 1%. In fact, I suspect the design will actually consume energy rather than deliver it. I'm not surprised they never built even a prototype, much less find a commercial application.

I actually like Gyorgy's use of the thermoelectric effect better. At least there you have a single step conversion from heat to electricity. NASA has been using this technique since the 1960s. In fact, Arthur C. Clark first suggested their use back when he patented the communication satellite in 1950. For example SNAP 3 was the first satellite to use an RTG,(Radioisotope Thermoelectric Generator), Pioneer 10 and 11, Voyager 1 and 2, and the New Horizons probe to Pluto all use RTGs. But, thermocouples are notoriously inefficient (5% is typical). Again, this could be expected because the temperature difference is quite low. But, if money is no option, and the energy source is large (ie: nuclear), and there's no other mechanism, a thermocouple design might be the way to go.

But, that brings up the economics issue again. One thermocouple isn't very expensive, but 10's of thousands, maybe millions is another story. Conventional construction techniques aren't likely to produce an economical solution. (NASA doesn't have to worry too much about economics.) On the other hand, nanotechnology offers some hope. At the moment, nano scale self assembly is in its infancy. But eventually, I expect the technology will mature to the point where it's possible to construct mol quantities (that's 10^23) of devices for little more than the cost of the materials. When this happens, we won't be just recovering waste heat from smokestacks, but everywhere there's a small temperature difference.

we can combine that with solar chimney but this become very expensive with low efficiency !!!!!