Wednesday, May 07, 2014

Sci-tech watch, 19: A look at Rankine-based cycle steam turbines (such as have been used in power plants and ships) and extensions -- for geothermal energy

The classic steam turbine power plant is an example of a system that uses a single working fluid that operates between two phases, e.g. water/ steam, in accord with the Rankine power cycle:

 The addition of superheat allows for greater efficiency, as it runs between a higher top temperature and the lower one, which inherently allows for higher efficiency. This is because the harvestable energy of a Carnot Cycle . . . the physically ideal (but impractical) heat engine . . . is at the physical limit set by the laws of physics:

Efficiency = (useful energy o/p) / (required energy i/p)

Input heat drives the cycle giving work and
exhaust heat at the cold temp
Carnot Cycle, running between Thot and Tcold:

Efficiency =  1 - (Tcold / Thot)
T is in Kelvins, relative to absolute zero, and so if Tc = 303 K (86 degrees F or 30 degrees C) and Th = 523 K ( 250 degrees C or 482 degrees F) then ideal maximum efficiency is 42%.

 In realistic cases, we may be at perhaps 1/2 that ideal, though some systems do better. 

250 degrees C is of course a reported temperature for fluids recently discovered in Montserrat through two exploratory GT wells that have been drilled down about a mile and a half underground.

For reference, Wiki notes:
In practice, a steam engine exhausting the steam to atmosphere will typically have an efficiency (including the boiler) in the range of 1-10%, but with the addition of a condenser and multiple expansion, and high steam pressure/temperature, it may be greatly improved, historically into the regime of 10-20%, and very rarely slightly higher.

A modern large electrical power station (producing several hundred megawatts of electrical output) with steam reheat, economizer etc. will achieve efficiency in the mid 40% range, with the most efficient units approaching 50% thermal efficiency. (Cf. onward reference to Power Engineering magazine, here.)

A classic video is helpful to understand how the turbine stage of the plant works:

 An important factor is that by condensing the steam, the back end pressure is reduced, allowing for a greater efficiency of harvesting energy from the flow of the working fluid. (This is an advantage over the windmill, which is unable to readily condense the downstream flow into a vacuum plus a liquid. That leads to the Betz limit where a wind turbine can extract only about up to 59.3% of the power in flowing air. A Steam turbine may approach 90% extraction from the steam flow, though of course a so-called back flow turbine is much simpler and cheaper.)

For geothermal energy, the source of the hot working fluid is a well bored deep into the earth which draws off hot fluids -- percolating rain-water or sea-water or both -- heated by hot rocks, which makes superheat a bit of a moot point, but eliminates the need for a fuel source for heating:

Latent heat and vapourisation for water etc (HT: WIki)
 Here, in a slide from Uni of Auckland, we see how perhaps about 3/4 of the fluid from the well is lost in the separator, and the 25% of harvestable steam is then used to run the turbine which turns the shaft that drives the generator. (NB: A significant part of the fluid may already be steam coming up from below, in a mixed phase fluid.)

Where, also, it must be recalled [cf. graph to the left], to flash one kg of water into steam in the relevant temperature range requires some 1.5 - 2.2 MJ of latent heat . . . depending on how close we are to the critical point of water (beyond which a liquid is not physically possible -- we have a permanent gas). That imposes a limit on the available steam and implies degradation of the heat content of the remaining hot liquid. 

So, the liquid remaining after flashing is significantly degraded, and there is a practical limit on how much of it can be flashed. 

The steam-flash approach generally requires temperatures in excess of  150 degrees Celsius in order to work adequately. Though, under reduced pressure, further flashing could be done. Such reduced pressure will however affect the available power in the flow of steam.

Below 150 degrees C, it begins to make sense to switch working fluids, to substances with lower boiling points, such as iso-butane or iso-pentane, which are used in an organic working fluid Rankine cycle, generally without super-heat. (Such cycles also make sense for G-T brines that are chemically aggressive, e.g. highly acid.)

That brings to bear binary G-T plants, e.g.:

As can be seen, the working fluid is flashed into gaseous state in a heat exchanger, the preheater/evaporator, which uses the hot G-T fluid (up to perhaps 150 degrees C) as heating source.

This causes an obvious additional loss of efficiency, which typically rides on the already low efficiency produced by having a low temperature source. [A Carnot cycle running between 150 degrees C ( 423 K) and the 30 degrees above, would max out at about 28%, with of course practical systems being much below this. That is, there is a difference between ideal potential and practical achievable output given losses and parasitic loads such as pumps, etc. Typical practical efficiencies for a binary G-T cycle may run 10 - 13%.]

That makes the Binary approach of limited use with small sources with high enough temperature, where whatever efficiency we can harvest will be at a premium. However, if we wish to wring out energy from the 3/4 or so of liquid that would perhaps otherwise be wasted, a binary module may be of help.  Auckland Uni provides a case, due to ORMAT:

Such modularity would also be of help in load following, as well as perhaps in switching from one unit to the next, in support of regular maintenance downtime.

Ormat says of such an approach:
Ormat Geothermal Combined Cycle Units (GCCU) are designed to generate power from high-pressure geothermal steam resources. When there are non-condensable gases present in the geothermal steam, our GCCU technology provides higher efficiency, lower maintenance costs and higher electrical output in comparison to conventional condensing steam power plants. Ormat offers GCCU modules in a wide range of sizes in order to accurately match the power plant to the geothermal energy resource. The result is significant savings in well drilling and resource development. The GCCU can be air-cooled so that all steam condensate can be 100% re-injected.  Total fluid re-injection sustains reservoir life and provides pressure support. Due to the low profile of the air-cooled condensers and the lack of any plume from the water cooling towers, the visual impact is minimized and the risk of acid rain is minimized as well.  
In turn, the use of turbines in all the possibilities above brings up the issue that turbines strongly prefer to work near rated load, pointing to the better load following capability of piston based engines -- never mind the disadvantages of such engines in terms of inherent tendency to more vibration/noise and maintenance requirements because of numbers of moving parts.

Of these the standard developed across C19 was the triple expansion steam engine, as is shown, courtesy Wiki.
Triple expansion piston engine

At each stage, a double-acting valve pushes the engine two ways, then feeds the next stage with its exhaust. As the steam will need to expand as its temperature is lowered, each stage is progressively bigger physically. (To give an idea of scale, the 1.9 MW triple expansion plants used in the old Liberty Ships of the Second World war to drive their screws were about 20 ft high and similarly long, and perhaps 7 ft wide.)

As Wikipedia comments: 

These engines use a series of cylinders of progressively increasing diameter. These cylinders are designed to divide the work into equal shares for each expansion stage. As with the double expansion engine, if space is at a premium, then two smaller cylinders may be used for the low pressure stage. Multiple expansion engines typically had the cylinders arranged inline, but various other formations were used. In the late 19th century, the Yarrow-Schlick-Tweedy balancing 'system' was used on some marine triple expansion engines. Y-S-T engines divided the low pressure expansion stages between two cylinders, one at each end of the engine. This allowed the crankshaft to be better balanced, resulting in a smoother, faster-responding engine which ran with less vibration. This made the 4-cylinder triple-expansion engine popular with large passenger liners (such as the Olympic class), but this was ultimately replaced by the virtually vibration-free turbine . . . 
Such devices, obviously are still a possibility, but they would be rather specialist items now, and the implications of vibration for wear, tear and metal fatigue should be noted.

Of course, for Montserrat, it is being suggested that the GT component should be joined to a diesel plant, and to wind etc as relatively minor components.

Diversity enhances reliability.

However, a major motivation for the move to G-T energy and other sources is to get out of dependence on oil, for economic and environmental reasons. 

Where, also, it bears noting that it is not unusual to see G-T plants turning in availabilities of 95 - 98%, placing them among the most reliable technologies in the world. That suggests that if the plant can be based on several modules of unequal size . . . say (for argument) 1. 8 MW flash, 1.2 MW Binary 1 and 0.6 MW Binary 2 [or, perhaps 3 x 0.6 Binary modules] . . . then we can see that the three units can be used in various combinations up to 3.6 MW that would allow significant load following and would regularly release each unit in turn for maintenance support, leading to a much enhanced system reliability.

Let us see how things come out in the end. END