Sign in

Libre University uses your GitHub account. Signing in is only needed to sit a final test, so the score is kept on your profile.

Superheat, reheat and regeneration

A basic steam cycle takes most of its heat in at temperatures far below its peak, and every improvement that has ever been made to one is an attempt to fix that.

One target

The previous lessons established the criterion. For a cycle rejecting heat at a constant TL, the thermal efficiency is exactly 1-TL/TH, where TH=qin/Δs is the mean temperature at which heat is added. The baseline plant, boiling at 3 MPa and superheating to 350 degrees Celsius with a condenser at 10 kPa, had TH=479 K and ηth=0.334.

There are only two levers. Lower TL, which means a colder condenser, and that is set by the river or the air and is already as low as the site allows. Or raise TH, which is entirely a design decision.

Three modifications do it, and the whole of this lesson is the same calculation performed four times. To isolate each effect the components are taken as isentropic throughout; real machines subtract about four points from every figure below, uniformly, as the previous lesson showed.

Superheating further

The easiest move is to keep the boiler pressure and take the steam hotter. At 3 MPa the tables give h=3457.2 kJ kg⁻¹ and s=7.2338 kJ kg⁻¹ K⁻¹ at 500 degrees Celsius, and h=3682.8 with s=7.5085 at 600.

Run the cycle at 600 degrees Celsius. The pump work is unchanged at 3.02 kJ kg⁻¹, giving h2=194.83 kJ kg⁻¹, so qin=3682.8-194.83=3488.0 kJ kg⁻¹. The exhaust quality is (7.5085-0.6492)/7.4996=0.9146, so h4=191.81+0.9146×2392.1=2379.7 kJ kg⁻¹ and the turbine gives 1303.1 kJ kg⁻¹. The efficiency is (1303.1-3.02)/3488.0=0.373.

Both defects improved together. The efficiency rose by nearly four points, and the exhaust moisture fell from 18.7 to 8.5 per cent, which is the more valuable of the two, because it is what limits blade life.

Check it against the criterion: TH=3488.0/(7.5085-0.6492)=508.5 K, and 1-318.96/508.5=0.373. Two hundred and fifty extra degrees of superheat moved the mean temperature by only 29 K. Superheating adds heat at high temperature, which helps, but it also adds a great deal of entropy, which dilutes the gain. Superheat is a moisture control first and an efficiency measure second.

The ceiling is metallurgical. Around 600 degrees Celsius, ferritic steels give out and austenitic or nickel alloys are needed, at several times the price. That is why almost every modern steam plant in the world sits between 540 and 620 degrees Celsius, and why raising it further is a materials research programme rather than a design choice.

Example. Run the same plant at 3 MPa and 500 degrees Celsius, condensing at 10 kPa, with h3=3457.2 and s3=7.2338. Find the efficiency, the exhaust quality and the mean temperature of heat addition.

The quality is (7.2338-0.6492)/7.4996=0.8780, so h4=191.81+0.8780×2392.1=2292.1 kJ kg⁻¹ and wturb=1165.1 kJ kg⁻¹. With qin=3457.2-194.83=3262.4 kJ kg⁻¹, the efficiency is (1165.1-3.02)/3262.4=0.356. The mean temperature is 3262.4/6.5846=495.5 K, and 1-318.96/495.5=0.356 confirms it.

Now you. Confirm the criterion for the 600 degree case: given qin=3488.0 kJ kg⁻¹ and s3=7.5085 kJ kg⁻¹ K⁻¹, find TH and the efficiency it predicts.

Answer

Δs=7.5085-0.6492=6.8593 kJ kg⁻¹ K⁻¹, so TH=3488.0/6.8593=508.5 K and η=1-318.96/508.5=0.373, matching the enthalpy calculation.

Raising the boiler pressure

The stronger lever is pressure, because it raises the temperature of the boiling itself, and boiling is where sixty per cent of the heat goes. At 3 MPa water boils at 233.85 degrees Celsius; at 15 MPa it boils at 342.16.

Take the cycle to 15 MPa and 600 degrees Celsius, where h3=3583.1 kJ kg⁻¹ and s3=6.6796 kJ kg⁻¹ K⁻¹. The pump work rises to 0.001010×14990=15.14 kJ kg⁻¹, so h2=206.95 kJ kg⁻¹ and qin=3376.2 kJ kg⁻¹. The turbine expands to 10 kPa at quality (6.6796-0.6492)/7.4996=0.8041, giving h4=2115.3 and wturb=1467.8 kJ kg⁻¹. The efficiency is (1467.8-15.14)/3376.2=0.430, and TH has climbed to 560 K.

Nearly six points from one change, which is why boiler pressures rose steadily through the twentieth century and why modern units run supercritical, above 22.06 MPa, where there is no boiling at all and the fluid passes continuously from liquid-like to vapour-like.

But look at the exhaust. Quality 0.804 means 19.6 per cent moisture, worse than the original 350 degree cycle and far outside what a turbine tolerates. The reason is geometric: raising the pressure at fixed temperature lowers the entropy at turbine inlet, so the vertical expansion line lands further to the left, deeper into the dome. Pressure and dryness pull against each other, and the whole point of the next modification is to have both.

Reheat

The fix is to expand in two stages with a return to the boiler in between. Steam leaves the high-pressure turbine at an intermediate pressure, goes back to a reheater in the boiler, is brought up to a high temperature again, and then expands through the low-pressure turbine to the condenser.

Take the 15 MPa, 600 degree cycle and reheat at 3 MPa. The high-pressure turbine expands isentropically from s=6.6796, which at 3 MPa lies between the tabulated 300 degree entry (h=2994.3, s=6.5412) and the 350 degree one (h=3116.1, s=6.7450). Interpolating, the fraction across is (6.6796-6.5412)/(6.7450-6.5412)=0.679, so the steam leaves the high-pressure turbine at 334 degrees Celsius with h=3077.0 kJ kg⁻¹, having delivered 3583.1-3077.0=506.1 kJ kg⁻¹.

Reheat at 3 MPa to 600 degrees Celsius, which takes 3682.8-3077.0=605.8 kJ kg⁻¹ of extra heat and returns the steam to h=3682.8, s=7.5085. The low-pressure turbine expands to 10 kPa at quality 0.9146, exactly as the 3 MPa superheated cycle did, giving h=2379.7 and a further 1303.1 kJ kg⁻¹ of work.

Now total up. Turbine work is 506.1+1303.1=1809.2 kJ kg⁻¹, pump work is 15.14, heat input is 3376.2+605.8=3982.0 kJ kg⁻¹, and

ηth=1809.2-15.143982.0=0.451

Both problems solved at once. The efficiency is two points above the 15 MPa cycle without reheat, and the exhaust moisture has fallen from 19.6 to 8.5 per cent. The mean temperature of heat addition is now 580 K.

The efficiency gain looks modest for the plumbing involved, and the moisture gain is what actually justifies it: reheat is what makes high boiler pressure usable at all. Large plants reheat once as standard and some reheat twice, with diminishing returns each time.

Example. The same plant reheats at 3 MPa but only to 500 degrees Celsius, where h=3457.2 and s=7.2338. Find the efficiency and the exhaust moisture.

The high-pressure turbine is unchanged, delivering 506.1 kJ kg⁻¹ and leaving the steam at 3077.0 kJ kg⁻¹. The reheat now supplies 3457.2-3077.0=380.2 kJ kg⁻¹, and the low-pressure turbine expands to quality 0.8780, giving h=2292.1 and work of 1165.1 kJ kg⁻¹. Total turbine work is 1671.2 kJ kg⁻¹ against a heat input of 3376.2+380.2=3756.3 kJ kg⁻¹, so ηth=(1671.2-15.14)/3756.3=0.441. The exhaust moisture is 12.2 per cent, still outside the usual limit.

Now you. For that same partial-reheat cycle, find the heat rejected in the condenser and check the efficiency from it.

Answer

qout=hexhaust-hf=2292.1-191.81=2100.3 kJ kg⁻¹, so η=1-2100.3/3756.3=0.441, agreeing with the work route.

Regeneration

Superheat and reheat both raise TH by adding heat at higher temperature. Regeneration does the opposite and better: it removes the lowest-temperature heat addition from the cycle entirely.

The complaint against the basic cycle was that a quarter of its heat went into warming feedwater from 46 degrees Celsius up to saturation, all of it at temperatures where it does the mean no good. Suppose instead that warming is done by steam bled from the turbine partway through its expansion. That steam has already produced some work, and the heat it gives up to the feedwater never passes through the boiler at all. The boiler then sees water that is already hot, and every kilojoule it supplies goes in at a high temperature.

The simplest hardware is an open feedwater heater, a mixing chamber at the extraction pressure. Condensate from the condenser is pumped up to that pressure, bled steam is mixed directly into it, and the mixture leaves as saturated liquid at the heater pressure, to be pumped again to boiler pressure. Two pumps, one vessel, no tubes.

Let y be the fraction of the turbine flow that is extracted. Per kilogram entering the turbine, the heater receives y kilograms of steam at hext and (1-y) kilograms of pumped condensate at ha, and delivers 1 kilogram of saturated liquid at hf. The energy balance is yhext+(1-y)ha=hf, so

y=hf-hahext-ha

Example. Fit the 15 MPa, 600 degree cycle with one open feedwater heater at 1.2 MPa, condensing at 10 kPa, with isentropic machines. At 1.2 MPa, hf=798.33 kJ kg⁻¹ and vf=0.001138 m³ kg⁻¹. On the expansion line at s=6.6796, interpolation between the 200 degree entry (h=2816.1, s=6.5909) and the 250 degree one (h=2935.6, s=6.8313) gives the extraction state. Find the extraction fraction and the thermal efficiency.

The interpolation fraction is (6.6796-6.5909)/(6.8313-6.5909)=0.369, so hext=2816.1+0.369×119.5=2860.2 kJ kg⁻¹. The condensate pump raises water from 10 kPa to 1.2 MPa, costing 0.001010×1190=1.20 kJ kg⁻¹ and giving ha=193.01 kJ kg⁻¹. So

y=798.33-193.012860.2-193.01=605.322667.19=0.227

The feed pump then raises saturated liquid from 1.2 MPa to 15 MPa, costing 0.001138×13800=15.70 kJ kg⁻¹, so the boiler receives water at 798.33+15.70=814.0 kJ kg⁻¹ and supplies qin=3583.1-814.0=2769.1 kJ kg⁻¹. Only the unextracted (1-y) reaches the condenser, at h=2115.3 kJ kg⁻¹, so qout=0.773×(2115.3-191.81)=1486.9 kJ kg⁻¹, and

ηth=1-1486.92769.1=0.463

Three and a half points above the reheat-free cycle at the same pressure, for one vessel and one extra pump.

Now you. Move the heater to 0.6 MPa, where hf=670.38 kJ kg⁻¹ and vf=0.001101 m³ kg⁻¹. The extraction state there is wet, at hext=2721.8 kJ kg⁻¹. Find y and the efficiency.

Answer

The condensate pump costs 0.001010×590=0.60 kJ kg⁻¹, so ha=192.41 kJ kg⁻¹ and y=(670.38-192.41)/(2721.8-192.41)=0.189. The feed pump costs 0.001101×14400=15.85 kJ kg⁻¹, so the boiler receives 686.2 kJ kg⁻¹ and supplies 2896.9. The condenser takes 0.811×1923.5=1560.0 kJ kg⁻¹, giving η=1-1560.0/2896.9=0.461. Almost identical: the optimum extraction pressure is broad, which is a mercy for anyone designing one.

What real plants actually do

The scheme above has one heater. Real stations have six to eight, at pressures spread between the condenser and the boiler, and the gain per heater falls off sharply after the first few, with the total worth about five points over an unregenerated cycle of the same conditions.

Most of those heaters are closed, not open: the bled steam condenses on the outside of tubes carrying the feedwater, so the two streams never mix and only one feed pump is needed for the whole train. Closed heaters are slightly less effective per stage, since the feedwater cannot be heated quite to the bled steam's saturation temperature, and they are much easier to plumb. Almost every plant uses one open heater, usually called the deaerator, because direct contact with steam is what strips dissolved oxygen out of the feedwater and stops the boiler corroding. Everything else in the train is closed.

Stacking all of this gives the modern ultra-supercritical unit: steam at 30 MPa and 600 degrees Celsius, double reheat, eight feedwater heaters, condensing at 4 kPa, reaching about 47 per cent net efficiency in service after auxiliary power and boiler losses. That is roughly the practical ceiling for a steam cycle, and it has been approached asymptotically for forty years.

The reason it is a ceiling is that every lever is now against a hard stop. The condenser sits at ambient. The peak temperature sits at what steel can hold. Regeneration has taken the low-temperature heat addition out. What remains is the boiler itself, where combustion gases at perhaps 1800 K hand their heat to steam at 600 K, and no amount of cycle rearrangement touches that, because it happens outside the working fluid.

Getting at it means using the high temperature directly, with a working fluid that goes through the combustion rather than sitting behind a tube wall. That is the gas turbine, and it is the next lesson.