MUC-L 40~ 20

I “This document consists of ... ‘5__
: pages and / figures
. ] =9 | o (NoesGe of  f4...copies, Scries..d
2 ndeddomgiioflfpes g LEETING OF JULY"6319%, & ”wh £ g sty
4. Coordinate With:

: ~9:00 |- 10:30 Eck - 209
6. Classified Information Bracketed

Reume |7, Other (specify):

o1 CHRRED gy

EE\SS\Y \oAT

Present: Messrs. Seitz, Hogness, Allison, ¥zfl ; YW Wér}gg{, Wein
Smyth, Cooper, Creutz and Ohlingen - //i?
ik -

Slehey
N /2/
Mr. Seitz presented a very interesting report on Riz @ sit; {he| Ceneral

FElectric Company at Schenectady to discuss turb: a-f?ys%'gms/afid in parficuls
new mercury turbine and mercury-vapcr processgs / ¢ A /
/ e SRS

This subject is of interest to u.'g( cause, first, the mercury"topping!

system is the most modern and efficient power mduqt?gffi"éyis and,
second, it presents samething new and dfifemn{*ifl%@gmfi on—-the uss of’
1iquid metal at high temperatures. The latter particular interest to us

because of' its application to certain of the potential pile designs involving the

use of liquid metal at high temperatures.

The mercury-vapor process is based on the fact that to increase tharmo-
dynamic efficiency, one should obtain the maximum possible workiag temperature or
rather the maximum possible differential temperature. The mercury system which
is a "topping" system would operate between two temperatures T; and T, to pro-
duce about 1/3 to 1/2 of the total power while the steam system would operate
between tempsratures Tp and T3 to produce the balance of the power, about 243
to 1/2, Most of the mercury-vapor systems in use now, have the temperature Tp
established according to the top temperature available in their existing steam
system, with the mercury system added to bring the top temperature up to Ty. The
engineers at Schenectady would much prefer to lower the value of Ty and recomrend
that new systems which are designed provide for a lower T in order to give a
better overall efficiency.

' The mercury-vapor process is a binary system for producing power from
fuel with groater thermal economy than is possible with the steam cycle =2lons.
The mercury cycle can also be considered as & steam producer in which, for a given
amount of fuel, nearly as much steam is produced as in a steam cycle and, in
addition, the by-product power from the mercury turbine generator is cbteined et
nearly the mechanical equivalent of the thermal energy. The advantags of the
mercury over the water system is that one obtains &s high temperature with merc
vepor as with water vapor at only 1/10 the pressure. One disadvantage of the
mercury system is that the weight and volume of mercury required is larger than
that of the water required in a water system. In a water system oporated at
about 9509F, the entropy per pound is 1.546 BTU/OF/1b and the volune of vaper par
pound is 0.3538, while in a mercury system operating at about 105C to 11C0°F,
the entropy is 0.1193 and the volume per pound is 0.3998. Trom this we see that
a mercury system requires about ten times the werking volume of a water systom.

The present msrcury units are designed to produce about 20,000 kw in the
mercury system alcne and one unit uses about 400,000 pounds of mercury. A%t the

. . R 3
— . -
AR -
- . » -
- v - B R
» ~ - M
~ .
. i f -
- . . . .
-~ ..
P .
- ‘
.- .
.
- ’
- A - .
I
. . - .
‘
o
K - . «

B v rea T i B e R e ey
YTREUTAL Ty
v s deld

S e mm . oAy e
i AT
eVt

v

TR QReat febd); e ' , RN
e CIRBLOT ORI CLEHO(5G ' ’

wi

!
l
l. o S ‘.~- £ pecisanysq
{
f

W st g B

e QUi TS “QLoausT P!

-
i
i
7

-
tor ot

et e yowmsos poEcregsme  § o ’
vhiomads e Lo ALE gEsRaeuinggag 107 : SR S
e T PoeEmptace eENes L o T T A
- - . . . 1% . - - B
o B850 527 Aramemeney (Gics j!zgzywps;}{e}gz DT
DELVELUTHL OF THEUSA DECIYR2MRICYLION SBAEM | -
»® --""“:‘f,,‘\‘ - . ':j‘ l ) . o T

present market cost of mercury of around $2,00 per pound, this means an outlay of
nearly $1,000,000 for mercury for one unit, The known world deposits of nmercury
are large and with a production no greater than the present top yearly production,
as much as 1,000,000 kw in cambined mercury-steam plants could be installed each
year. Whereas these binary units might be installed on large ships, at present
they do not look very practical for smaller power requirements of & mobile natwre
such as trains, etc.

Although the figure above shows the utilizaticn of 20 pounds of msreury
per kw, the eagineers at Schenectady believe that a minimom of 8.25 pounds of
mercury/kw can ultimately be achieved. They have set up as a basis of compariscn
of efficiencies a figure based on the total overall efficiency frem the heat con-
tained in the coal to the power at the busbar. On this basis, the maximum theoret-
ical efficiency which can be cbtainad according to the first law of thermodynamics
is 3,413 BIU/hr/¥kw. The best overall efficisncy cbtained to date in any of the
mercury-steam cambination units is 9,175 BTU/br/kw. The best overall efficiasncy
obtained from an all steam unit is 10,039 BTU/hr/iw. This means the 21l steam
power plant has an efficiency of about 33% compared to the first law efficisncy
noted above while the binary system has an efficiency of about 37% or 4% gain
over the all steam plant., The operating temperatures for a typical binary unit
are: 1,050°F for the mercury or Ty (this is saturation temparatqre) and 925°F
for ths water or Tp (this includes the superheat). Thus the thermodynamic
efficiency ccuputed from the temperatures or the second law efficiency would
be about 65%.

The drawing attached illustrates diagremmatically the mercury-vepor
mocess for the production of power. Referring to the drawing, mercury is vap~
orized in a boiler at comparatively low pressure and passed through a mercury
turbine which drives a gemsrator. The vapor from the turbine is exhausted to a
condenser boiler where its latent heat is transferred to water vhich vaporizes at
any desired pressure, The steam formed in the condenser boiler is superheated in
coils located in the gas passages of the mercury boiler and is then used in steam
turbines or for process work.

The top limit to the temperature Ty at present is set by the tubass in
the fire box. These are actually located in the hot coal geses immediately above
the fire bed and have temperatures on the outside surface of around 1200°F and
on the inside of around 1075°F. Under these operating conditions, it is obvicus
that the tubes are easily attacked by the gases and slags in the fire box and
must have high creep strength as well as stainless properties. The best alloys
found to date for these tubes are the Sicromo series having about 1% each of
silicon, chromium and molybdenum with the balance iron. ihen bsttor alloys are
found, Ty can be raised above 1075°F and the efficiency improved. However, in
our type of power production units, we do not require a furnace and so We can
better the velue of Ty,

P T
Keagosirriadt
-

POAASNY

Mercury, as the material for the liquid metal portion of this binary
system, is a good material for this purpose because it dissolves very few metals
used in commercial high temperature practices, In fact, the solubility of all
metals in mercury is less than one part in 10[*. One problem, however, with the
use of mercury was the difficulty in obtaining good wetting of the boiler tubes
by the msrcury. This is especially necessary in the fire box and condenser
boiler to obtain good heat transfer, It was finally found that ty adding 0.5 ppm
of titanium and 10 ppm of magnesium as wetting agents, good heat transfer coef-
ficients could ba realized. These wetting agenis are probably efiective because
they are both good oxypen getters, taking it away fram the mercury and ths ircn.
Mercury probably tends to wear awey the iron axide while the titenium and mag- *
nesium, in taking the oxygen away from the mercury and iron, help the mercwry
in cleaning off the scale and keep the tubes clean and easy to wet. When starte
ing up a unit, it takes about ten minutes before wetting occurs but from that time
on, no further trouble is encountered if about one pound psr month of titaniun and
possibly a very small amount per year of magnesium is zdded. Unfortunately, a
small amount of air (about 1 cu ft per hr) leaks into tue mercury turkine at the
condensing or low pressure end. This prcbably accounts for the nacessity for ths
extra titanium.

In an all steam power plant, an entire corpa of chemists is required for
constant water analysis and treating to avoid scaling while the mercury system is
practically self-meintaining except for the small addition of titanium. Thereiore,
in addition to the increased efficiency, the mercury system offers a tremendous
reduction in maintenance costs,

One great worry in a mercury system is the pessibility of leaks in view
of ‘the high mercury costs. In the many operating years credited to mercury-vapcr
binary systems, only one leak has occurred. This was brought about as a result
of the following process: powdered coal was fed into the top of the fire box
while ashes produced thsrefrom fell into the bottom of the furnacs and were swept
out by streams of water. In the case of the single leak, water splashed up and
caused steam which attacked the tubes with a resulting loss of about one ton of
mercury through the leak.

Despite ell premonitions, the mercury turbinss have caused no troubls
and have operated very successfully without pitting or erosion. The careful da-
sign of the turbine to produce impact incidence of the mercury with tho turbine
blades at a safe angle instead of perpendicular appears to be the secret for the
successful erosion resistance of these turbines, They are made with high carbon
steel parts instead of stainless steel as employed in steam turbines. However,
at present, the steam turbine is more efficient than bhe mercury surbine based
on the overall transfer of kinetic energy in the gas to mechanical energy from
the turbine. For & steam turbine the efficiency is about 85% as compared to
only 73% for the mercury turbine. The latter figure is lower probably bscause
of the tendency of the mercury to condense in the turbine.

The operating pressure for the mercury system is around 125 pounds per s in, and
for the water system around 1,250 pounds per sg. in. et

SR

To consider the application of this information to our problem of de-
signing a new high temperature pile, we must considor the possibility of a metallic
alloy of uranium or plutonium which can be used molten through a pile which would
replace the fire box in the above process. There are only two systems giving =
eutectlc with uranium (and presumably plutonium) with Low enough melting points to
be practical. These are the iron-uranium (or plutonium) alloys and the nickele
uranium (or plutonium) alloys. The latter eutectic melts below 75G°C and has
about 40 atomic percent nickel, The difficulty with the first named eutectic vis
that iron dissolves everything which can possibly be uscd for the tubes.

Very little is known about the mercury-urenium alloys which are pyro-
phoric. kore data should be obtained on the phase diagrams of this slloy end
further information obtained on the nickel-uranium alloys, Of course, the pile
need not be operated with the active metal in molten form circulating through the
pile but may have another metal in liguid form as the coclant with the uranium or
enriched material stationary within the moderator. An example of this type is
My, Szilard's bismuth cooled pile,

Another problem in using a urenium eutectic is that the fission pro=-
ducts will be contaminating the liquid metal constently. At Hanford, this nay take
only a few hours. While most of the fission products are good elameébs and might

help as wetting agents, icdine and the alkaline metals might interfere with the
cleansing action. g .

iy, Ohlinger questioned ths supposed reduction in
ste Mr. Seitz agresd that there was no roduction
n m% the mercwy cystem has been added simply
xisting steam plant but that any newly desigaed unit

e reduced, it is to be expected that maintenance

~discussion period with a brief discourse cn a
: , in which the hot geses would replace the coal
gase 3 X ¢ ‘binary system, the active metal would be locatad
hroughout the moderator in lumps either of urenium carbide or molten wranium con-
ad. - ,_of bery oxide. The furnace in the binary system
: ger for absorbing the heat carricd off
Cooper's objaction to this scheme was
be necessary in order to get sufficient
around 10 atmospheres). lMr. \einbarg's
exceedingly large because of the high tem—
do. In fact, he felt that it would be
~cooled grophite-moderated plant. Mr. Hog-
that would be used up in recirculeting
of helium.
-5 -

At subsequent meetings, Mr., Wigner will discuss his "pulsating” pile
and those utilizing endothermic chemical reactions and Mr. Szilard will discuss
his seed pilas, '

Jdp
STEAN y MERCURY VAPOR

SUPERHEATER —_—

(_J (gh 4 —
[ _

U

d—— -
j[— T e —i )
Nie= DA AR
(C— ‘U‘WW
\ = —— L x_ |
\
\\ { ‘ MERCURY GENERATOR
L) Fummace Juuuuuls. TURBINE
) = -— MERCURY CONDENSER
| MERCURY BOILER (é:—-b -
MERCURY FEED PUMP
STEAM
‘ STEAM TURDINE GENERATOR
: ( «
,i TAAA AN .
WATER
A

_~SOTEAM CONDENSER

—fi

J)
WATER FEED PUMP -

F"Om"ffuflaacfact‘/on_fl of tte /f.f,/f,f.f,'" CED-YC4