V’..-

4 U. S, NAVAL TECHNICAL MISSION IN EUROIE c/o Fleet Post Office New York, N.Y.

File: (40/Hn)

Serial? 1443 13 October 1945 r.*

From Chief, U.S. Naval Technical Mission in Europe To Chief of Naval Operations (0P-16-PT).

Subject: U.S. Naval Technical Mission in Europe Technical Report No, 257"^5, Nitroglycerin, Diothylono Glycol .1 Dinitrato and Similar Explosive Oils - 'fcnufacturo and Development in Germany « Forwarding of.

Enclosures: (A) (HIV) Ono (1) oomplotu copy of subject report as ■1 listed in distribution. (B) (l£7) Ono (1) sot of negatives of photographs in * subject report with copy No. 1 as listed in distribution.

Enclosures (A) and (B) aro forwirdod herewith

2. CJO (OP-16-PT) is requested to raako oomplotu addit­ ional conics Of this report for forwarding to suoh other agencies as may bo interested.

HOI’SINGrR ’’ Captain, U. S. H Acting •1 DISTRIBUTION To?

CNO (OP-16-?1?) v/Enalosurc B Copy No. 1 MA?...... Copy No. 2 .1 Offloe of Rosoaroh and Invention ... Copy No. 3 CamNavEu •••••..••*,... Copy No. 4 CoaWavi’.u for Admiralty ...... Copy No. 5 BuOrd (Ro) ...... Copy No. BuOrd (Ro 2a) • •...... Copy No. BuOrd (Rr 6) ...... Copy No, 8 BuOrd (Pr 6a) . . . . . Copy No. 9 Naval Powder Factory, Indianhoad, lid Copy No, 10 NavTecKisSu (itain Files) ...... Copy No r

TECHNICAL REPORT NO, 257-45

NITROGLYCERIN, DIETHYLENX GLYCOL DIN1TRATE AND SIHI^R

EXPLOSIVE OILS - MAWACTURS AND DEVELOPMENT IK GERMANY,

SUMMARY

This report describes the manufncture end development wo He on explosive oils nt three representative German plants. Two types of contihuous units and the batch precess equipment were observed.

Aooosslon Tor nhs~”gra&i SEPTEMBER <945 DTIC TAR s ■Un'innouin’r-^

By.

Avnilr^ll*tv Codos U. S. NAVAL TECHNICAL MISSION IN EUROPE and/or Diet

'Cr<0

RESTRICTED

NITROGLYCERIN, DIETHYLENE GLYCOL DINITRATE AND SIMILAR

EXPLOSIVE OILS - HAIJUEACTURE AND DEVELOPMENT IN GERMANY.

1. Xfttn>4iMUga» Plants for ths manufacture of the explosive oils wore investlgai- teA by this team at Schlebusch on 38 Jone, at Krummel on 15-July, and, at Bonllts on July -17th and. 21st. -

I

1 The Dynamit AG riant at Schlebusch had three units, the old batch process, a modified Schmidt-Melssnor continuous plant and the Biazsi unit, the most modem, This plant made only Nitroglycerin (NG) and Ethylene Giycbl Dinitrate (EGN) for use in ocnaexclal dynamites nade there.

The DAG plant at Krummel had four batch nrocess units, two for the dynamite oils and two for the Diethylene Glycol Blnitrate (DEGN) and similar oils used for cannon and rocket powders.

The Elba plant of Wolff and Co, at Bomlits had six complete Schmitt- Meissner continuous units, producing only DEGN for powders. There were five small unite and ono l?rge one.

German cannon and rocket powder compositions were developed.pri­ marily taking advantage of the desirable properties and. the availabil­ ity of Diethylene Glycol Dinltrnte (DEGN), The manufacturing process and equipment for this material were based on tho experience gained in the plenta making Nitroglycerin (NG) nnd Ethylene Glycol Dinitrate (EGN) for dynamite.

On the three plants visited, Information was obtained on three types of nitration equipment which cover the most important differ­ ences in modern technique.

Until about 1925, all dynamite 'olants used the batch process for nitration'of NG end EGN. At about that time Schmidt developed a con­ tinuous nitration and neutralizing unit at Krummel works of Ttynamit AG, the first dynamite factory cf Alfred Nobel, the inventor, A num­ ber c«f those units were built at other plants but there are none now nt Krummel, Serieus accidents occurred at several of the installations. •j

RESTRICTED

IitotogUsa (QgnWt)

Because of the inherent hazard of the Schmidt type of NO waste acid separator, the Biazzl Co, of Italy developed a continuous unit, one of which was observed on operation at the Schlebusch Works of JUG, After its successful operation was demonstrated, the Schmidt unit at that plant wns changed by replacement of the separator with a Biassi type edparator,

The ESGN units observed at Bomlits were the stmdnrd Schmidt typo as manufactured by the Meissner Co, The staff at that plant readily agreed that the Biassi separator was superior and preferable to the Schmidt separator. They preferred, however, the Schmidt wash­ ing system to the Biazzl.

This report covers only the Information on explosive oils and reference to their further processing and use is contained in the fol*4 lowing technical reports!

Nitrocellulose and Paste, No. 258—45, Cannon Powder, No. 25G-45, Rocket Powder, No, 260-45.

3,

The NG niirator units at Schlebusch were built in bunkers with a horizontal cylindrical concrete roof covered with earth. One end of the room was made of frame and glass windows to sei^e as a blow-out wall in case of explosion. In all three units, nitration, reparation, water washing, and soda neutralization were carried out in the one room.

Tor nitration of Glycerin or Ethylene Glycoldinitrate and mix­ tures of these, the Schlebusch dynamite plant had three units. The first and oldest wr.s the Batch process. The second was a Schmidt- Meissner, modified by use of a Biazzl separator. The third w°s the most modem, the Biazzl unit. •

(a) Batch Process,

The equipment used for the batch process and the sequence of operations is shown in the attached diagram.

V. V

J

RESTRICTED JrtUfriart. -. (a) (font1 d.) ’

The nitrator charge was 200 kg of Glycerin end 990 kg of nixed .’i*. acid containing 51-52> S?03 and 51-52$ BjSO^.

The nitrator was made of Y3A alloy, the separator was lead, and the wash tanks were made of VaA,

The tine cyclo for nitration was one hour. As the equipment was so snail and had no interesting features, other details were not recor­ ded, Tull detailo on the DAG equipment and process for batch nitration are described in the Krummel part of this report, _• > ■J (b) SchmidtrKel^ner Process. The details of the Schmidt-Meissner unit at Schlebusch have not been recorded because complete description of this equintaent is given in Part IV - Bomlitx. The original separator and washer towers . *1 *j ’4-v on this unit had been replaced by Blassi type separators and washers because of the difficulty of cleaning and the< inherent hasards. In most aspects the Schmidt unit was then quite similar to the Biassi,

(c) Blast! Process, t-.v: The equipment sequence and slr.es for this process are shown in the attached diagram. A picture of the nitrator is also shown. All of it was located in the one house cxcent tho final W«G. separator and the wash water setting tanks. ir*»r**i In the Blaxxi process, small vessels (50 liters) with intensive mechanical agitation were used for washing the NG, These mixtures were then separated continuously in larger vessels (360 liters) which were quite similar to a conventional cyclone separator. All of the equip­ ment was B’de of V2A alloy. xhe welding of this equipment appeared to be very well done. r<4

The Glycerine or Ethylene Glycol mixture with Glycerin was fed ■i continuously from a constant level tank through a Rotameter tjujpe flow­ meter. The mixed acid was added through similar equipment, xhe anal­ ysis of the mixed acid was 51-52$ HNO3 end 51-52$ HgSO^. The mixed ». acid to Glycerin ratio.was 4,9 parts to 1 part. ! The capacity of tho unit won 800 to 1000 kg, (2200 lbs.) of NG y

►

BIa33I n.‘«M

wic -ilH T4 i

CtuLING _

SEPARATOR 3oC L

RECOVERED -* N.G.

XASHER ASTER \ !“« 50 L SESAR TOR W.A. j I > WASTE ACID HITRATING HOUSE SEPARATOR 360 L SEPARATING house *• WASTE n'ATER wM VrtSH

WASHER NUgCOj SGLUTIO( L->. 50 1

* RECOVERED N.G h WASTE rfATER WASHER > 50 1 N.G. WATER SEPARATOR SEPARATOR 360 1 360 L r?j;is

TO N.GtBUGGY

iA’AT. '

RRSTHICT7H> .fchlofrtfth (eXfr>nV

nor hour.. With a volume of 200 liters In .the nitrntor, it was stated thrt there was about 80 kg of KG retained in tho vessel.

The nitrntor hod six cooling colls. The agitator blade turned at 400 HPK and pulled the charge downward. The degree of emulsifica­ tion of the oil and.aoid was highly important since-tn© much agitation- hindered proper generation. *ho temperature of tho charge was main­ tained at 23«>25n 0r '

■ Tho nitrntim ni.V.uri was run from en nutlet pipe at tho top of the nitrator over to tho aldo, inlet of the separator. -This inlet was-* placed to provide tangential flow. The separator was a cylindrical tank with a conical bottom* *ith the eaulsifled-feod fed in at the . middle-point, the separated waste acid.was run out the bottom, through a trap rising un to a line to the kfter separator. The. acid oil was run off from a top outlet of the separator and over to the first waslr- er. ®he simple construction and the lack of baffles or pockets in tho 3iaxsi separator were distinct advantages over tho Schmidt separators,- It was stated that separation was unite uniformly good and no agents such as guhr for improving senaratlon were noedod. .f;

The first washer for tho acid oil was a 50 liter agitated vei?sol where the WB«h«wntcr was thoroughly mixed (430 with the oil con- , tinuously. ^he overflow of tho mixture was led to the next separator which was built similar tp the other vessel described* above. Trom hero tha wash water was sent to a final separator in tho. next building and then either to the sewer or to the*denitratIng recovery tower. -Any <*11 collected in tho final water separator wos returned to the system at tho neutralizer washer.

, T^o.pre washed oil was next run to two agitated washers in Ser** les. xhese had been' provided tn allow the addition of podium carbon­ ate solutipn at two points,. In practice, the soda solution was added only in -the first. The additional agitation provided in the second mixer ansieted in obtaining complete neutralisation of the oil. The emulsion of oil and so‘da water 'was the? run by gravity thru V2A pipe • to the second building, about 50 feet iway, fnr final sop. ration. '

Thia building was ."built bo a bunker similar to the first but .with the blowout 'wall at right angles to it. The final separator for oil and-soda water wps exactly the, sane as the others described above. 'c

•. • J

aSTJSCTsp Schlebneeh j(e)(Cont’d.)

‘4 The separated oil was run out the trap and through a line to a third building where it was collected in a buggy; or a tank on wheels, for transfer to storage tanks.

Th* soda wash water was mh to another separator and then to the ditch, lay oil recovered here was returned to the system,

The standard operating procedure required daily draining of the whole system. ®he nitrator and.separator were than flushed out tho­ roughly with a synthetic wan to acid (mixed acid of the snrce composi­ tion but no dissolved NG) to remove all MG, Duo to tho simple unob­ structed passages and piping in the system, very gnnd donning wns obtained, Tho femo.inder of the equipment ws washed by running water through it. At shut down, the synthetic waste odd remained in the first two vessels and the rest of the equlnctent wns filled with water.

The unit was started the next day by addition of the Glycerin and mixed acid, first at reduced rata and then at normal rate, The agittv* tlon soeeds and the cooling capacity, once established, required very little further adjustment.

It pas stated that a yield of 334 lbs, of NG was obtained from 100 lbs« of glycerin — n conventional yield, 4

Ths ^lassi unit was completed in 1943, Airing «;he first six months of operation some small alterations to piping were made but . since then the Schlebusch staff has considered this process and equip­ ment to be most satisfactory*

Soon after the start of operations, there occurred one stwill ac- cl dent when a slight explosion ripped open c. piece of the >11816 acid line from the first depuratnr to the after separator, ^hls occurred on a Sundry afternoon while tho unit’was.shut down and damage was neg^ ligible. It was due to a small trap in the line which hnd not washed clean. This line was straightened and no further accidents have oc­ curred. . * -.1 r." The above information was obtained from Dr,. Demoff, Acting Plant ?«• Manager, and Dr, Von Sommerfeld the NG Aren Superintendent, The lat- ter is considered by other German MG technicians to be one of the

'•’.I present leading authorities in this field, While at Schlebusch a vis-

• ■

•A I V v:

i£hi Schlebuich (c)(C^td.)

It was paid to Dr. Ph. Naoum, .forcer head of J^G' 8 laboratory, now re* tired. Ke was the author of the World-renowned. hook on Nitroglycerin and la considered tn ho one of‘the world's authorities on high explo­ sive manufacture. Dr. Naoum has not been active since 1939 when he lost his leg in an explosion of an experimental unit. I 3. Xrursnel". :

A very thorough description of the XruJnmol NG operations was ire* pared by tho staff at that plant. A translation of this report is ns follows! • • '•Technical and Manufacturing Processes of Digly-coldinitrate -• Triglycoldinitrate and Hethrioltrinitrate.

The explosive oils necessary for the production of powder paste were made in Xrunael in 2 plants. In the older plant (during the last few years only Triglycoldinitrate was aade here) nitration and washing wee conducted in separate buildings, In the new plant the complete process from nitratinn to washing was carried out in nno building. The new plant had two nitmtion houses* Here Di glycoldinit rate was made exclusively. Tor the Nethrloltyinitrato a trial plant was Installed in one nitration house. Production on a large scale such as in the case of Diglycoldinitrate had not yet been carried out.

k The production of ^iglycoldinitreie from diglycol was by menhe of esterification with nitric acid according to the following equation!

CEL — OH CH3 ~ ONOo 1 1 . ' CH3 — 0Ha

y) 4 2

CH2 . CHja 1 1 CHgOH OH2*ONOa (04H|003) (O^OyNg)

Since' esterification is an equilibrium reaction, for nitration a pure nitric acid was .not used but a mixture of 65{C HNOg and 35^ H3f04. » •

-10-

W ■ I ■T y» i.’

HZSTRI Krumol (Coat’d.)

The sulfuric acid combined with the resulting water and thus shifted the balance of the reaction toward the deter aide.

In Krummel only the discontinuous or batch process vrs employed. Charges of 420 diglycol were used. The average yield nf Diglycold- inltrr.te was ca. 710-715 kg per charge) i.e., in round figures 170{C. Theoretically 777 kg Diglycoldinitratc should be obtained from 420 kg diglycol, Part of the loss of, explosive oil was in tho vnste ncid, and also d scalier part was lost by washing,

Dlglycol is a liquid clear as water and has a soeci'.c gravity of 1,126. As a rule it contained ca. 6.14 moisture and up to 14 glycol. The nitration was carried out by slow addition of dlglycol to the ni­ tration acid with intensive air stirring and good cooling with brine. The latter was a 354 sodium nitrate solution which was numped at a temperature of -15° through the cooling nockets. *

As already mentioned, a mixed acid was used having the following composition: ,

'•N 65^ HNOj .35{C H2S04

The acid should be as free from nitroso ns possible. It was ob­ tained either ready mixed from the 1.0. or else mixed in Krummel from 264 Tuning sulfuric acid and 98^994 nitric acid or HS acid which con­ tained 88*» HNO3 ano HgSO., The initial ratio* was 112,9 i.e. for 420 kg diglycol 1318 kg mixed acid was used. Of that ca. 790 kg was HHO3 and 428 kg HySO.a Theoretically only 497 kg HKO3 are necessaxy, N Thus an excess of 393 kg was used. By this relatively high excess a higher stability of the resulting waste acid was obtained. Therein lies the great difference between the waste acid of diglycol dinltrate and nitroglycerin. While the KG acid lasts for days and could bo ’’stored* in the aft or-separation* vessels the dlglycol waste acid had to bo processed immediately because it had a tendency after several hours to decompose. At the same tine the temperature rose rapidly and large quantities of nitrous gases escaped, emanating from the oven­ , heated ester.

The maximum temperature of nitration was 25° 0. Two thermometers I r»«- each with a different dip-depth mnde it possible to watch over the tem­ perature in the nitration apparatus, ^he nitration with a cooling

-11- hi

V £01

IMTgCTa J&aaislJ&nlhL}. • brine of -14° to -15° lasted ca. 23-35 minutes. After nitration the explosive oil acid mixture was cooled down to'lS0 and then drained in­ to a separator,- Here the heavier waste acid nenarnt.id t acid oil. The separation lasted 7 minutes. Tho clearly nenaratod waste acid was drained off for denitration, tho acid oil run into the preliminary washer in which threre were oa. 300 liter water, and a wash water consisting of ca. 340 HNOj, 50 HgSO^, 3-30 ester, remainder water resulted. This washwater was also used in the denitration. The waste acid just montioned had the following composition: *

64-660 HnSOx 8-9 0 HHO3 • 4-5 0 ester, rest water. , • Trom the preliminary washer the explosive oil then wag run into the final washing nnperatus. There first, washing with ca. 500 liters cold water was used. Then with 150 liters ca. 50 soda solution, heated to 60°. The soda wash lasted 10 minutes with Intensive air-stirring, The final washing was mode with ca. 500 liters cold water, Prom this wash­ ed” oil a sample wns token to the laboratory, ^he test (reaction to potassium lodide-starch paper) must last at 83° at least £0 minutes, " If the t-'st wns satisfactory, the oil wns then released for the manu­ facture of paste.

The Diglycoldlnltrate is a yellowish oil with a spec, gravity of 1.38 - 1.39 and a nitrogen content of 14.1 - 14.20, The deconroosltions point is 1629. At 300*’ it explodes. Sensitivity to impact by a 2 kg drop hammer amounts to 160 cm. Diglycoldinitrate is in this respect, compared to nitroglycerin (with 2 kg drophnmmor only 4 cm), consider­ ably less .sensitive. The freezing point is below -10°, The ^ater sol­ ubility is ca', 0,40. ' • • • . As to the nitration plant itself, all buildings wore massive rein­ forced concrete structures which were surrounded by earth barricades. The front" side of these buildings had a large window meant bo be a blow-out wall. In tho case of an explosion it would, fall. toward the outside 'and thus would moke on escape route for the high pressure, 3ach building had two escape corridors, one at the upper exit and one nt the lower exit. In the building were the nitration apparatus, separator, preliminary wesher and final washer, all made from V2A. The required cocks were of clay, and partly armor-plated. The drowning

. ‘* -12- A

BgSTHICTSD

Krummel (Cont*d.) r, tank (safety vat) was’of'iron lined inside with rubber. It was largo enough to receive the whole charge from the nitration or separation apparatus with a goodly amount of dilution with water, As the name indicates, it wm installed in order to prevent the charge from ex­ ploding in case docomposltioh sots in by means of strong dilution. If during nitration or even, afterwards an abnormal rise in tempera^ turo was noted, then the operator 'whould immediately “null the emer­ gency lover. In this way, the automatic safety installation was put in action. 3y means of the tho operating nir which was present in the air mains— I Beil signals wore switched on by way of a relay in and on top of tho building, also In the adjacent buildings, • (2) By way of a so-called gun (kenone) the cocks of the nitra- tion apparatus and the separator were opened towards the drowning tank,

(3) A membrane valve was opened, permitting fresh water to flow, for the purpose of dilution, into the safety vat, and,

(4) The air-stirring in thb separator and safety vat was started. In case of disturbances in the plant nir each nitrating house had a reserve air vessbl which could be switched over in ox* c. der to finish the charge which was In progress.

Communication between the houses which belong together was carried out by means of signal installations, e.g., between the weighing house and tho nitrating house, between the nitrating Muse and nil storage, between nitrating house1 and denitration. Only, after confirmation of a U^it signal by a return signal could operations in question be resumed.

The capacity of the entire nitration plants for regular operations wan stated ns 60 tons per day or 1500 per month, on tho basis of con- Ml . tinuous 3 shifts andbnsed on 25 working days. Tor this 25 operators nnd 1 foreman were required.

In the nitration of t’riglycol a nixed acid composed of 70$ HNOg and 30$ HgSO^ was used* The initial ratio was It2i5, l.e., for 500 kg

,4 -13-

M

h

& •V-VA

Hg3TRICTBD fossa?! .(fcnVdJ. * . • • trlglyool 1350 kg nixed acid w,i used. The yield ^mounted to roughly 650’kg with the above mentioned initial mixture. xhe solubility of the triglycol dlnitrate in the waste acid was very high, amounting to on. 8-9$. ^he donitrntii.n of this waste acid was difficult on account of the high explosive oil content. Inasmuch as the separation which last­ ed 30 minutes and more, was subject to certain dangers the’ oil-ncld mix­ ture was diluted after nitration and this acid water km then processed tn the denitration* In this manner the acid losses remained within the usual*limits, , • . t The manufacture of the triglycol dlnltrAte proceeded according to the following equation: <

OBnOH I3 0H20H03 C^3 C«3 ^0 0

OHa V 4 2 HjO • cHa °?3

** 0 0

OHg

CiigOH CH20N0a ' Wi!

■ Basically the manufacture of trlglycbl dlnitrate was the' same as .that of diglycol dlnitrate, with this difference:

Nitration time amounted to roughly 30-minutes. After 3 washings with water a soda wash with 30 kg soda per initial mixture takes place, thereupon the usual aft er-wash. * e • • The triglycol dlnitrate .is a brownish oily liquid with a speci- . Tic gravity of 1.335. xhe nitrogen content is 12.1—12,2$, Theoret­ ically it should bp only?11,7$. ®his difference is explained by’the diglycol content of the trigiycol amounting up to 3$. The sensitive-

-14— >A-1 • • • •

RESTRICTS!)

Krummel (Cont^d.)

nogs to impact of the triglycni dlnitratp is still less than that of dl- glycol dlnitmta. A test with the. 2 kg fall hammer showed, no effect. The iodide starch naper test should last 20 minutes at 82 decrees.

Tho capacity of’the nitration ulnnts ft> r the. manufacture of triply^ col dinitrate was stated to be 44 .tons per dsy.

In addition'to dlgiyeol dinitrate and triglyool dlnltrpte oooae- tonally smaller amounts of methriol trinitrate were also manufactured. Tor this nitratioh the same mixed acid was used as for diglycol dini­ trate, thus

65$ 35$

Tho initial mixture ratio was 1:3,5. The formation of explosive oil takes place according to the following equation:

z OHgOH , 0£>0H02 %-o- OHgOH 4 3 IM03 HgC—OHgONO, ^CHgOH !• s OHgONO* Ws1 ’’sWs1 The average yield was ca. 200$. The nitrogen content of the explosive oil theoretically amounts to 16.47$, the actual values were between 16.32 - 16.36$.

While the methriol is a white, uowdery material with a moisture • content of ca. 0,5 - 1$, the ester obtained from it is an oily, slight­ ly cloudy liquid whose .density is 1.460 at 20 degrees. The decomposi- tlon of the explosive oil starts at 182°. The sensitivity to impact corresponds to that of nitroglycerin: 2 kg fall hammer 4 cm height. The oil is practically Insoluble in water. The starch iodide paper test should last 20 minutes at 82°.

A small experimental apparatus was-available for the nitration of methriol. ^or the xsirpose of introducing (into tho, anparatus) methriol which le a fine powder (previously sieved), an apportioning screw was

-15-

r- ~ «i KB1TBICTZD

Krummel (Cont«d.) » t installed which permitted, the material toxbe fed in a fine 'stream as evenly as possible. The formation of lumps sfibuld bo avoided tit all costs, for they may easily cause decomuositlon, • • » The period of nitration for 50 kg methriol- was ca. 30 minutes pt a maximum brine temperature of 30°, Tor the separation ca. 15 minutes should bo required. The temperaturb in the- pre-,washer and at the so­ da wash should be 40° because at lower temuerature methriol is consid­ erably more visc.nus than diglycol dinitrate. The soda wash lasts 30, minutes, thereupon the usual aft er-wash with water.

The cauaclty of the ojroerljiental plant for methriol trinitrate vp.8 60 tons per month.

The new plant for*the manufacture of explosive oil differs from the old plent mainly in that the nitration and washing wore done in one building. Thus a considerable labor saving wns obtained* While the old plant required 4 mon por shift, the how ">lnnt with 2 nitration houses, required only 5 men. A dlsp.dvmtr.go, however, was that accord­ ing to safety regulations the nitration of a new charge could bo start - ed only when the finished oil of the previous charge had left the build­ ing. This regulation was changed so that between 3 operations there had. to be 2 empty vessels. Practically, the result wrs that a new ni- • tration was started at a time when the previous one wns in the soda wash..

The completely washed explosive oil was then transferred to the storage. The* explosive oil line was located in a. tunnel in whldh a man could walk. At the end of the tunnel wns a swinging arm from which the dll could be run into a wagon. "This wagon was large enough to take a. whdle charge, i.e,, ca. TOO kg exnlosive' oil. ®he wagon-was used to transport the oil into tho real storage. Here, according tn regula­ tions, up to 4000 kg explosive oil could be stored. The storage was built so that it had 2 ■exactly equal” halves, tach for’2000 kg explo­ sive oil. From each half 3 paste mixing houses could be supplied with explosive oil. In each half there were 3 rectangular tanks tach with 700 kg content.. When the explosive' nil was rdn ’out df the Magon it flowed across a filter which lyid to be washed regularly'. At the bot-' tom of the tanks there were drainage cocks from which the oil could be drained off into a scale by wey of a. rigid pipeline and a swinging arm. This scale was used to weigh the ex-nlosive oil per charge for each in-

’ ' -16- isstriotsd

Krummel (Cont{d,) dividual initial fixture of paste. The explosive oil flowed from the scale into a container fron which it was ejected by means of water into *•*

(a) Denitration of the Sxulosivb Oil!-Waste Acid. C’’.’

In the denitration operation of the factory at Krummel tho wa^te acid emerging fron the' separator, of the nitration house and the wash acid, which was formed by drowtlag of the acid oil, were mode > usable again. • . .

Tho acids were freed from tho explosive oil dissolved in them by <£• running through the so-called breakdown columns **t high temperature and thus rode usable again. In the case of the waste acid from diglycol ’1?.^ normally uo separation of the exnloslv'o oil took place, or else ohly in [ traces. ; ‘

Due to the action of moisture the .waste acid fron diglycol was easily decomposed bnd therefore could .only be stored for a few hours at the most. The decomposition started at one place and spread farther ’ and farther nt a Jump-like increase of ^emnomture. Tho liouid then boiled and developed largo quantities of nitrous gases. •<**••* • The same process occurred in the breakdown columns, however, here it was regulated and continuous. In order to increase tho oxygen sur- •-'X\ jfllus already .present in tho nitric acid in the waste acid, and thus to be sure that all ester was burned, a 50$ nitric ncid or, in its place, _ wash acid wns added. /. << ■'

. Incomplete burning of the ester wns shown by dark coloring of the wnsto sulfuric acid es & result of suspended carbon particles. Incom- nlote oxidation could enuse a clogging of the columns by means of-or- gnnic material. . ‘ j «

The ester dissolved in the wash acid, wns destroyed merely by^run- ning"through the breakdown column and by addition of steam at 100° 0.

‘•X1 h’ 'X‘X' 1 * . . . t__ -17- * ■ - ■ . cXX C.-.’.i MSTRIOTZP

Kami .(q) (coattAiL * .

Operating Rules.

At tho otmrt: Careful hooting of the column. , 4 hr. 75 cbm water 150 cbm H SO^/lO sec. | hr. 125 cbm * 350 " 2* H t hr. 350 cbm * 400 * , " * then addition of steen up to 4 gradation marks of the scale. ■

, 1» Denitration nf th? Waste Acid, ■

*' 1000 « 1500 cbm wnste acid • 80 cbm wash r.cld/10 sec.

3. Temoernture of the Column:1

nt the top 130 • . at tho bottom 160 , Steam until necessary temperature is reached;

3. Working of Wash Acid:. *

800 ~ 1200 ebe wrnh acld/10 sec. Stenm until necessary temperature la reached. • Temnornture pt upper end of column 100° . , Temperature at lower end. of column 105°

4. Concentration of Nitric Acjd, *

(Wns carried out only when the building was available.) 800 cbir. sulfuric acid/10 sec. 500 r.bta nitric acid/10 sec. Addition steam . Result: 96$ nitric acid.

Physical data p.nd a nummary of the re-worked and npnufaaturcdacld.

Waste acid from diglycol (not yet denitrated)

One batch: 385 liter , ■ Spec. Wgt. 1.66 ’ • •

-18- BSSTHICTZD

Kruanol (nXCoht'd.)

Vnit« mH from dislycol (not yet dcnitr-sted) 4*50 ••ter dissolved in the acid 64-660 ILSO, 90HHO3^ 4

W« A^jd, (denitrated)

Sj»oc. wgt. 1.64 715? h2S0, 0* »304

VMh Acid (Prior to running through the denitrator!

,Ono hatch: 400 liter Spec. wgt. 1.36 3-30 ester 60 33504 33-340 HBO3

Vaah Acid

Tower: 45-500 HN03 • Bleaching apparatus 38-400 HN03 Tank: receipt from tower and wash ncid from denitrator 330 HN03 3.30 H2S04

Bgaarha

The donltmted. waete’acid was tested, with Fe304 solution during op­ erations for the presence of HH03, or N304r respectively.

The diglyonl waste ncid from the nitration house flowed into n stor­ age container and from here it ran hy wny of the DAG separator into the denitrator. The,ester dissolved in the waste acid was burned by addi­ tion of wash acid or nitric ncid and steam. The nitrous gases thus de­ veloped were drawn by n. fan..t hrough the .condensers end forced into the absorption tower. Back of each condenser* a-blcwbtag apparatus was con­ nected in which the nitric acid, fl.owing off from there was freed by means of air-suction from the nitrous acid contained therein. The nit­ ric acid ran by way of a rotameter tn the tank.

-19- BMTBICTED

Krummel (ft)(Cnnt,d.)

The sulfuric acid fran the lower end. of the donltratnr rnn through a siphon into tho cooling column, from thefo by way of a cooler into nn intormodicry vessel and then wee uunpcd into tho tank. The air cur­ rent emanating from tho cooling column still containing smoko partic­ les pulled along frnn the sulfuric acid wns, prior to its release into tho open, conducted into n condensing column whldh wrr surayed with wa­ ter obtained from the condensers.

In the pcncontrati n of nitric acid tho ncid went to the blench­ ing apparatus by way of n nre-wnrmor heated with steam into the exhaust column whore tho nitric oxides present were removed by air induced by n fan, Tho acid ran through a siphon into an intermediate container and was pumped from there into the tmk.. Tho wash acid ran from the star- . are container by wey of » small separator and a nitrocellulose filter into the .denitrator, Tran the denitrotornthe acid flowed through the cooler into tho intermediate container, xhe cooling columns were not, . needed hero,

• At the Bnnlits plant of Blbla, thoro were six Schmidt-Meissner ni­ tration units installed to rapke SBGM for tho paste which was used for ennnon and rocket nowder, Tho five arall units each had a cnuaclty of 300 kg per hour and the one large uhlt had. a capacity of 1000 kg per hour, « • Tho MG area was described by Dr. Werner Schmedding, Superintendent of the HG. and Acid- Area operations. He had been nt tho plnnt for six years and .had supervised the start up operations in 1939. There had been no explosions or serious accidents nn this area, *

Dr, Schmedding w>.s well acquainted with many other explosive oil । factories in Germany and is urobnbly the best authority on Schmidt-Meis­ sner units. ■

After Schmidt developed this equipment, the manufacture and instal­ lation of plants was undertaken by the larae European firm rf Meissner and Go. ®his concern trained thb first plant operators but did not pro­ vide operating manuals. Unfortunately, the plant drawings of this equip­ ment could- not be located. ‘

It was Dr. Schmedding’s opinion that the best continuous iTG equip-

-20- B3STRICTSD

Booliti (Cont'd.)

nent would consist of Schmidt nitrators, Biar>zi scparatora sad Schmidt wnshers.

loch unit consisted of three bunkers or earth barricaded concrete igloos with one end node of frame. In the first, there were two rooms, one each for Dlethylene Glycol (DEG) tanks and mixed acid storage tanks The second contained the nitmtor, separator, and washing systems. ®he third contained the NG storage tonka end.the vessels for emulsifying the oil nnd water for,transfer to the Paste nixing houses. I * fl

The expenditure for construction of this plant' was apparently much more than that on the DAG plant so The provisions for camouflage trees K» nnd for sinking the buildings into side hills wexvj much more obvious.

The raw material, DEG*was obtained in tank cars from the I.G. fac­ tories nt Huls/Heckllnghnusen, near Essen and at Leunn/kerseberg near Berlin. Each car lo*d was analysed for water content, for unstfturates J and the boiling noint range was determined. The maximum specification for water was 0.5$ as determined by bensene distillation, xha specifi­ cation for unlaturatee allowed none as determined by the test with sil­ ver nitrate and ammonia. ®he‘boiling point range was 246—349° C.

Thu nixed acid used had the following composition:

63$ MSO, 36$ ►wrrtl

A ratio of 3.1 parts mixed ncid to 1 part DEG was used and a waste* acid of the following composition wae obtained:

1056 ’• hso3 * .

BjO

It was also necessary to provide a synthetic waste ncid with the following composition for cleaning the nitrntor nnd sonarntors 1* 10$ hnd3 • 60$ H3SO4 . 30$ NgO

-21-

El *,• *•*

MMTHCgp BoJXU*.&aVdJ

. It WI stated that «, yield, of 173 lbs. of MGN ym obtained from 100 lbs. of MO when using the most efficient wash water recovexy eye­ tea. Tho overall plant Nitric Acid loss wee 0,75 Ibe. per lb, of MON sade, Tho theoretical consumption would be 0,65 lbs. of nitric acid.

The raw materials, MG nnd nixed acid were fad to the nitrator by means of air pressure on the storage tanks, through RotMieter tyne flow­ meters. The mto wns controlled by the air pressure valves.

The large nitrator unit only 4.8 described in the following para­ graphs, The snail units were nf identical design but of smaller dimen­ sions.

Tho MG was introduced thru a vertical nine which extended to c, nolnt 3 inches above tho 1 inch hole in tho glass cover. No distrib­ utor was needed.

The nixed acid was introduced at the bottom of the Hitrntor.

The nitratnr was made of VIA alloy in three main parts flanged and bolted together. It was about 30w diameter nnd 40* high. The ton sec­ tion was cylindrical, about 10 inches high, and wns covered by a loose glass nlate allowing very gond observation. The center section contain­ ed tho cooling element. This was a set of tubes open at the ton and bottom through tube shoots, ^he dooling brine (NaltOs solution) was cir­ culated in the chamber surrounding tho tubes. In the center of the tube sheet there woe an opening of about 8 inches for the agitator shaft. The bottoa section was conical with a 4 inch outlet on the bottoa.

Agitation was provided by a propeller and paddle at the point just below the bottom tube sheet nnd punned the nitrator charge up the cen­ ter opening over the top sheet nnd down the tubes.

At the bottom outlet there was a quick opening drowning valve and also an alternate drowning means consisting of a Mipolom plastic disc which could be cut open by a knife actuated by a temperature controller and from a switch outside the house. .

The agitator shaft was driven from a variable speed, carbon brush motor located in r room walled off from the nitrator room proper.

—2^*

r- S' V

L fe RESTRICTED

The motor electric drive was connected to safety controls which shut off the nitrntor feeds in case of electric failure. On the motor shaft there was a flywheel 18 inches in diameter to provide agitation for iha time interval between electric failure and the Introduction*of the last DEG.

£ It was stared that the proper agitation speed had to be determined for onph individual nitrator finite Kost of them varied between 380 .to 400 RPM. The determination of the optimum speed w^s made by mixing. Wa­ ter and paraffin oil stirred at 30° 0. Upon stopping agitation, this mixture should settle to n clear line in 10 minutes.

In starting up a Schmidt unit, tho nitrator and’separator first filled with synthetic waste acid. The mixed ncid was then start­ ed nt 5/6 rote. After 30 seconds the DXO was started nt the sane rate. At exactly 7 minutes later, n clear separating line should be observed in the separator sight glass. If the line hod not formed, the nw materials wore immediately shut off and the synthetic waste ocid was Started to flush out the system before starting again. If nonaal sep­ aration occurred after 7 minutes the nites were increased to the stan­ dard.

In normal practice the nitrntor charge was maintained at 30° Ct The volume of the charge was 350 liters.

The Schmidt separator consisted of three main prrts. Tha lower two wore cylindrical vessels flanged together and supported at an angle of 45° from the horizontal. 2he upper section was a vortical ellipti­ cal t*>nk about 15 inches high on the straight side with the ton tapered to to a 6 inch dinmeter outlet. The bottom was flnnged to the top of tho middle sedtion. ' In operation, tho line of separation was about 8 inches nbnvo the flenge. , * a The mixture of DEGN nnd waste acid overflowing from the nitrntor was run into the separator at a noint at the "bottom of the upper sec­ tion. The feed pipe 'branched to three perforated nipes inside the sep­ arator to provide for distribution, A sight glase in the front of the uuper section allowed observation of the charge to determine the level h® of the interface of oil and acids.

The middle section contained ulateg parallel to the eloping walls,

-33- M BESTRjsra

i*e., nt a 45° nngle* These plates were provided for th© stated purpose of allowing the drops of separated oil to rise to the top section while the waste acid settled to the bottom, .The numerous plates spaced about J inch apart made cleaning difficult.

The bottom section of the separator was a cooler constructed with tubes and sheets similar to that described for‘the nitrator, Brine was circulated thru the chest of the cooler to keep the waste acid nt a low temperature and assure maximum separation efficiency.

The waste ncid flowed from the bottom outlet, un a riser and thru a sight box tn nn outside storage tank. The height of this riser and the height of the oil dmwoff the ton.could be adjusted for uroper operating levels of the separator.

The separator wns equinnod with automatic drowning valves and pits- tic diets similar to those described for the nitrator,

Trom the separator the acid bil was run by gravity to the first of four glass washing towers, These were made of 17 sections of glass pipe 20 cm in diameter nnd about 8 cm *hlgh, Between each glass section there was a sieve plate, Plastic gaskets and aluminum bands were used to seal the Joints, • I Into the bottom of the first tower, air nnd water (from the second tower) were introduced through a Jet nossle to nix them thoroughly with the oil run in through a third pipe. This mixture flowed up through the slave plntes"which provided distribution nnd agitation. At the top ov- erflow the>alr was, vented nnd the oil water emulsion was run into a rec­ tangular V2A box separator, •i • • * Prom this separator the oil was run down to the bottom inlet of ■ the second towor. In‘ this tower- fresh water,, precooled, was added. After the separator at the top, this water was run through a cooler and I. * -1 then was used ue the feed water for the first tower described above.

By using the waver in series in the first two towers. Dr, Schmed­ ding stated that the yield of DEGN was improved to 173 from 170, the yield with regular Schmidt operation using fresh water in both towers. Also a wash water was obtained which contained 46$ HNO,j and 3$ DBOT, This was then ’tent to the denitrating recovery. • • • * • • » * - • • • *

Bqnllti (

The oil from the second tower separator was then run through the last two towers counter current to the 5$ sodium carbonate solution used for neutralisation. This neutralizing solution'was separated and lint• to the sewer. * . . “ >> •’ Iron the lost neutralizing tower separator the oil was run to storage tanks in an adjoining room, in the bunker. , $ere it wns emul­ sified with water nnd jetted .to the large Paste mixer. For the small nitrator and Paste mixers, the oil was weighed out as the mixer charge required, ennlodf 5. s'. as described. In the Krummel section,, and jetted to the mixer house. • Xn addition to theqwquipment described, there were auxiliary tanks in each nitrator mom for (1) holding .synthetic waste acid, (2) the war- ter filled drowning tank, (3) neutralizing by batches, the nil separa­

ted in the shut down procedure.• » ,.• . » t . * • t •Prepared by; C.Hi BRQOKS, ' ‘ It. Cmdr., USKR.

C.X. REUSS, ‘ . Lieut., USSR. ‘

’ T.l. MILLER, ■ • Lieut., USNR.

. _ ' Dr. V.E. UWSOH, . . ’ ’ . ’ . . Technician?. ' • ' ‘ • ■ . CASTN3R, # Technician. ’

I

• -25-