Archive Record

  • Email This Page
  • Send Feedback
Title Article: The Eighth Wonder: The Holland Vehicular Tunnel. Gray & Hagen. Smithsonian Annual Report, 1931.
Object Name Article
Catalog Number 2011.005.0064
MULTIMEDIA LINKS CLICK HERE to see the PDF, text transcription and all pages; note - please be patient while the file opens.

CLICK HERE to see RTF, text only; note - please be patient while the file opens.
Collection Holland Tunnel Collection
Credit Museum Collections. Gift of a friend of the Museum.
Scope & Content Article: The Eighth Wonder: The Holland Vehicular Tunnel. By Carl C. Gray and H.F. Hagen. As published in Annual Report of the Board of Regents of the Smithsonian Institution; Washington: GPO, 1931. pp. 577-607 followed by Plates 1- 30. Four illustrations/maps. As received: removed from publication. Text and plate captions in notes. Text file plus text/image pdf on file.

This monograph appeared in the Annual Report as "Reprinted by permission, with a few omissions from a pamphlet entitled "The Eighth Wonder," published by the B. F. Sturtevant Co." It is considered a very useful document regarding this significant engineering achievement.

First published in 1927 (in book format) by the company that made the ventilation fans for the tunnel. Author Hagen was a fan engineer and was associated with the Sturtevant Company. The work is essentially the same, but in the original the plates were throughout the work.

Notes 2011.005.0064

CONDITION OF THE INSTITUTION FOR THE YEAR ENDING JUNE 30, 1930 (Originally published in separate edition, hard bound, by B.F. Sturdevant Co. in 1928.)

By CARL C. GRAY and H. F. HAGEN (2)
[With 30 plates]

Back in the second century B. C., a certain Antipater of Sidon
composed an epigram in which he enumerated what he termed the
" Seven Wonders of the World." They were the walls of Babylon,
the statute at Olympia by Phidias, the hanging gardens at Babylon,
the Colossus of Rhodes, the pyramids of Egypt, the mausoleum at
Halicarnassus, and the temple of Artemis at Ephesus.
To-day any similar list of wonders, no matter by whom compiled,
would doubtless include the pyramids, not merely because they alone
have survived the ravages of time, but because they still represent a
marvelous achievement of man's handiwork. What the other won-
ders would be might afford material for a contest sponsored by some
newspaper columnist. But surely there would be a place in such a
list' for the Holland Tunnel, as the longest subaqueous tunnel in the
world, a stupendous project, magnificently conceived and executed.
And surely old Antipater himself, however wedded he might be to
his own wonders, would to-day be glad to add the Holland Tunnel
to his list, as an eighth wonder of the world.
It is with this belief that the following record of its history has
been written, in recognition of the magnitude of the task, of the
heroism of its first chief engineer, Clifford M. Holland, and his
successor, Milton H. Freeman, both of whom gave their lives to the
undertaking, and of the great advance in the science of ventilation
which its construction made possible.
Of course, a tunnel is no new thing. Primitive man, living close
to nature, could hardly have failed to observe evidences of tunneling
by animal life about him, and soon made tunnels for his own pur-
poses. We know that in ancient Egypt a king, upon ascending the
throne, began at once to excavate the long narrow passage leading
1Reprinted by permission, with a few omissions from a pamphlet entitled " The Eighth
Wonder," published by the B. F. Sturtevant Co.
2Grateful acknowledgment is made for valuable data obtained from the official reports
of the New York and New Jersey Tunnel Commissions and from the Engineering News
Record, and for permission to reprint a portion of an article from the magazine Charm,
published by L. Bamberger & Co., Newark, N. J.


to the rock-hewn chamber at Thebes which was to be his tomb. From
Egypt, too, comes the first record of a subaqueous tunnel constructed

[Illustration] Figure 1. Plan and profile of the Holland Tunnel

under the dry bed of the river Euphrates, which had been tem-
porarily diverted from its channel. It was 12 feet wide, 15 feet high,
and was lined with brick masonry.


In the time of Caesar Augustus, or perhaps even earlier, the Ro-
mans built a notable tunnel through the Posilipo hills, between
Naples and Pozzuoli, about 3,000 feet long and 25 feet wide. In
order to light this tunnel, its floor and roof were made to converge
gradually from the ends to the middle: at the entrances it was 75
feet high. The Romans were the greatest tunnel builders of an-
tiquity. During the Middle Ages tunnel building was chiefly for
military purposes. Every great castle had its private underground
passage from the central tower or keep to some distant concealed
place, through which to make sorties, receive supplies, or escape in
time of need.
With the advent of gunpowder and of canal construction, a strong
impetus was given to tunnel building in its more modern aspect of
commercial or public utility. Previous to 1800, canal tunnels were all
through rock or hard ground. Then, in 1803, a soft-ground tunnel 24
feet wide was excavated for the Saint Augustine Canal in France.
Timbers were laid to support the roof and walls as fast as the earth
was removed, and the masonry lining built closely following. From
this experience the various systems of soft-ground tunneling since
employed have developed.
The use of shield and metal lining marks the greatest development
in the art of soft-ground submarine tunneling. The shield was in-
vented and first used by Sir Marc Isambard Brunei in excavating the
first tunnel under the river Thames at London, begun in 1825 and
opened in 1843. In 1869 Peter William Barlow used an iron lining
in connection with a shield in driving the second tunnel under the
Thames at London.
The modern tunnel shield is a steel-plate cylinder whose forward
edge acts as a cutting edge. Its rear end, extending backward, over-
laps the tunnel lining of cast-iron rings. Inside the shield, hydraulic
jacks act against the tunnel lining as a thrust block so as to push the
shield ahead when pressure is applied. A partition prevents earth
from entering the shield except as permitted through suitable open-
ings. As the shield moves forward, the lining is erected under the
protection of its rear. In submarine tunneling compressed air
pumped into the forward end of the tunnel counterbalances the pres-
sure of the water which tries to enter.
In 1906 the Legislatures of the States of New York and New Jersey
created for each State a bridge commission to investigate the feasi-
bility of constructing a bridge over the Hudson River, uniting New
York City with Jersey City. Legislative recognition was thus given
to an increasingly vital problem-some means to supplement the
ferries plying between these two ports.
Further legislation, enacted from time to time, continued the life
of these commissions. In 1913 they were authorized to consider the
possibility of a vehicular tunnel. Finally, on April 10, 1919, author-


ity was granted them to proceed with the construction of a tunnel,
or tunnels, between a point in the vicinity of Canal Street on the
island of Manhattan and a point in Jersey City.
Those who had the project closest at heart felt that the tunnel
1.Shorten the time of transit across the Hudson River and afford a con-
tinuous means of communication between New York and New Jersey, unaffected
by climatic or other interference.
2.Relieve traffic congestion, already serious.
3.Accelerate the movement of necessary supplies into the city of New York,
and thereby relieve conditions of distress.
4.Increase the tax value of real property within a considerable radius of
the tunnel terminals.
5.Pay its cost three times over within 20 years.
6.Reduce the high cost of living by reducing the cost of trucking.
7.Increase the facilities for commerce in the port of New York by removing
from the surface of the harbor many lighters and other floating equipment.
8.Furnish means for the uninterrupted movements of troops and supplies
to and from the city of New York in case of need.

The commission selected as chief engineer, Mr. Clifford M. Holland,
tunnel engineer of the Public Service Commission, First District,
State of New York, in immediate charge of the construction of
all subway tunnels under the East River. He was regarded as
having had a greater and more successful experience in the work of
subaqueous tunnel construction than any other member of his pro-
fession. A board of consulting engineers was appointed, and a
contract or treaty between the two States was drawn up and
approved by the commissions and given the consent of Congress.
Chief Engineer Holland took office on July 1, 1919, and at once
began the organization of an engineering staff. His chief assistants
were selected from those who had been associated with him in the
construction of the East River subway tunnels. Having had not
less than 10 years' experience in subaqueous tunneling, they were
well qualified both by technical training and by practical experience
to meet the requirements of the work. Actual construction began
October 12, 1920.
Upon the death of Mr. Holland on October 27, 1924, at Battle
Creek Sanitarium, where he had gone in search of health after devoting
all his strength and energy to the construction of the tunnel,
the commissions gave it his name. Under his direction all the more
difficult portions had been completed and the remaining details
planned, and on the very day his body was borne to his home there
came a demonstration of his engineering skill and accuracy in the
successful junction of the under-river headings of the north tunnel.
His successor, Mr. Milton H. Freeman, had been his division en-
gineer. He, too, gave himself unsparingly to the work, and died on
March 24, 1925. He was succeeded by Mr. Ole Singstad, who had

been engineer of designs under both Mr. Holland and Mr. Freeman.
Under his direction the Holland Tunnel has been completed.
The Holland Tunnel is located in the vicinity of Canal Street,
New York City, because that street is a wide east and west thorough-
fare giving direct communication across the island of Manhattan.
On the east, Canal Street connects with the East River bridges and
Brooklyn; on the west, with the Hudson River water front, at ap-
proximately the center of down-town traffic over the Hudson ferries.
Its location in Jersey City is at the logical point as nearly opposite
Canal Street as is practicable, in order to obtain the shortest tunnel.
This point is very near the center of traffic and is advantageously
located. It gives direct communication to Jersey City Heights and
points beyond by means of the Thirteenth Street viaduct. The water
front, with important railroad yards, is easily accessible and ade-
quate communication is afforded with the low-lying parts of Jersey
City and Hoboken through streets which parallel the river.
The southerly tube for eastbound traffic extends from Provost and
Twelfth Streets, Jersey City, under the Erie Railroad yards, the
Hudson River, and Canal Street to Varick Street, New York City.
The northerly tube for westbound traffic extends from Broome Street
midway between Varick and Hudson Streets in New York City,
curving to the west to Spring and Hudson Streets and under Hud-
son Street and the Hudson River, the Erie, and the Delaware, Lack-
awanna and Western Railroad yards to Fourteenth Street at Prevost
Street, Jersey City.
In planning a public undertaking of the magnitude of the Holland
Tunnel, consideration had to be given to many features besides those
of actual tunneling. The building of the structure itself was a
great engineering problem, but many investigations beyond mere
technical design were required.
To secure the best location and arrangement of tunnel roadways,
a survey of present and future traffic and the influence of the tunnel
on the development of adjacent territory was called for, first of all.
Traffic conditions had to be considered from many angles, such as
capacity, congestion of the tunnel roadway, adequate approaches,
congestion in adjoining streets, width of roadway, and the growth
and development of vehicular traffic.
A preliminary forecast of tunnel traffic, based chiefly on the yearly
increase in traffic over the Hudson ferries, resulted in an estimate of
the number of vehicles that would use the tunnel as follows:
1924 (when tunnel was expected to be opened) 5,610, 000
1935 13, 800,000
1937 15, 700, 000
1943 22,300, 000


Further estimates indicated that a 1-line tunnel would have a
capacity about equal to the traffic demand at the opening of the
tunnel. A 2-line tunnel would have sufficient capacity to accommo-
date all traffic up to 1937, while a 3-line tunnel would reach its capacity
in 1943.
Obviously it would be unwise to construct a 1-line tunnel whose
capacity would be reached as soon as put in operation. As between
a 2-line and a 3-line tunnel, it was found that the difference in cost,
with interest, would be sufficient to pay for another 2-line tunnel
after the first 2-line tunnel had outgrown its capacity. Of greater
importance was the consideration that no street or section could
accommodate the volume of traffic represented by a 3-line tunnel.
If a 3-line tunnel were built, it could be operated at only 2-line
capacity. This would violate two of the main principles governing
proper tunnel planning-the distribution of traffic so as to avoid
undue congestion, and the investment of capital for construction only
as facilities are needed, without the necessity of providing for the
distant future. These are two of the most important features in
which tunnel construction is held to be superior to bridge construc-
tion in crossing wide, navigable rivers.
The cost of a long-span bridge does not vary directly with the
span but increases about as the square of the span. On such a bridge
no commensurate saving in the cost of construction is obtained by
omitting some of its facilities. The tendency in bridge construction,
therefore, is to provide facilities greatly in excess of immediate
requirements, with a consequent expenditure of capital long before
those facilities are needed. Then when there is sufficient traffic to
utilize the bridge to full capacity, the resulting congestion in the
vicinity of the bridge entrances becomes a serious matter. This is
seen in the case of the East River bridges in New York City to-day.
Tunnel construction, on the other hand, is more flexible than
bridge construction, because the cost is a direct function of its length,
with the volume of excavation increasing as the square of the
diameter. Since the cost of excavation represents a large part of
the total cost of a tunnel, any increase in the width of roadway can
be made only at considerable expense. The proper way to plan a
tunnel is to avoid the disadvantages inherent in bridge construction,
build only for the present and near future, and construct other
tunnels at other locations when the facilities of the first tunnel are
Since a 2-line tunnel would have sufficient capacity to accommo-
date traffic up to 1937, and a 3-line tunnel would create such traffic
congestion in the vicinity of its entrances and exits as to preclude its
use to capacity; also since the difference in cost between a 2-line and
a 3-line tunnel, with interest, would pay for a new 2-line tunnel

when the first was outgrown, the obvious proceeding was to construct
a 2-line tunnel and when its capacity is reached, to build another
2-line tunnel at some other location as determined by future traffic
conditions. The Holland Tunnel is, therefore, a twin-tube tunnel,
providing in each tube for two lines of traffic in each direction.
In planning the entrances and exits of the tunnel, a careful study
was made of vehicular traffic, with particular reference to its move-
ment at street intersections and through the tunnel. It was recog-
nized that wherever traffic intersects, its continuity is broken. In-
stead of moving in a steady stream, it breaks into a series of waves
as it is held up and released at intersections. This interruption in
the stream of traffic at street intersections so limits the capacity of a
street that its real capacity as determined by its width is never
A tunnel differs from a street in that the only interruptions by
cross traffic are at the entrances and exits. Consequently these
points are of vital importance, affecting as they do the ultimate ca-
pacity of the tunnel. Unless the entrances and exits insure con-
tinuity of traffic during the period of maximum demand, the
capacity of the tunnel roadway can never be reached.
Accordingly, the entrances and exits of the Holland Tunnel are
widely separated. In New York City, one is to the north and the
other to the south of the Canal Street through traffic; in addition
they are located so as to be served by two main north and south
avenues. Tunnel traffic is thus given the best possible facility for
free movement while at the same time the greatest separation is se-
cured at a reasonable cost. In accord with this same principle the
entrance and exit at the Jersey City end are located in separate
streets adjacent to the railroad yards east of the north and south
traffic streets connecting Jersey City with Hoboken.
This separation of the tunnel entrance and exit traffic is consid-
ered to be a factor of the greatest importance in relieving congestion
in the vicinity of the tunnel. This was particularly necessary in
New York City, with its large and rapidly increasing volume of
traffic. It was also called for in Jersey City, where there were no
wide thoroughfares in the vicinity of the tunnel.
In addition, property was taken to provide broad plazas at en-
trances and exits. The entrance plazas serve to accommodate the
waves of traffic as they approach the tunnel and converge in the
portal roadway into continuous lines of vehicles through the tunnel.
Similarly wide exit plazas insure the free and uninterrupted move-
ment of traffic away from the tunnel. Through the separation of
entrance from exit, and the use of adequate plazas, the tunnel traffic
can be distributed over a large number of streets.

In considering the requirements for the width of the roadways
and the clear headroom needed, measurements were taken of vehicles
crossing the Hudson on the ferries between New York and New
Jersey. It was found that their height varied from 6 feet 6 inches
for passenger cars to a maximum of 13 feet for large loaded trucks,
but that the number exceeding 12 feet in height was only 1 per cent.
The width of motor vehicles varied from 6 feet for passenger cars
and light trucks to a maximum of 10 feet 6 inches for army transport
trucks. In the case of 3-horse teams, the outside dimension of the
three horses abreast was 9 feet, but the number of vehicles exceeding
8 feet in width was only 3% per cent.
In determining the amount of clear headroom required, it was
necessary to consider the matter of providing sufficient area in the
tunnel roadway. Any increase in clear headroom, without increas-
ing the size of the tunnel, could be made only at the expense of the
available ventilating duct area. Any reduction in this area would
increase the power required for ventilation and add to the cost of
operating the tunnel.
Given a maximum height of 12 feet 2 inches and a maximum width
of 8 feet, a clear headroom of 13 feet 6 inches seemed adequate to
allow even for jacking up vehicles in case of breakdown, and this
was decided upon.
Normal operating conditions in a tunnel accommodating two
lines of vehicles in the same direction on one roadway obtain when
there is a slow line of heavy trucks 8 feet wide abreast of a fast line
of light trucks and passenger cars 6 feet wide. It is, however,
necessary to provide for such a contingency as when a vehicle of
maximum width has to pass another of the same width that has
stalled. The roadway has to be sufficiently wide to permit the pas-
sage abreast of two vehicles of maximum width.
It was believed that in the slow line, operating at a speed varying
from 3 to 6 miles per hour, a clearance of not less than 6 inches
between the outside of the tire and the curb should be provided. In
the fast line, due to the greater speed, this clearance should not be
less than 1 foot. It was also considered that for safe and convenient
operation a clearance between moving vehicles of 2 feet 9 inches
should be allowed. These considerations led to the adoption of a
width of roadway of 20 feet, with, in addition, a sidewalk 2 feet
wide in each tunnel. This sidewalk is set back from the curb line
a distance of 6 inches and is located at an elevation of 26 inches
above the roadway.
This roadway is paved with granite blocks laid in the usual sand
cement cushion layer, about 1 inch thick, with the joints filled with
hot asphalt mixed with heated sand. By means of squeegees, a thin
coating, sprinkled with sand, is left upon the surface, resulting in a

smooth, resilient, and long-wearing surface that will help to deaden
the sounds due to traffic, and be more quickly repaired than concrete.
Each side of the roadway is lined with a granite curb, the roadway
having a transverse slope from one side to the other, with a de-
pressed concrete gutter behind the curbstone on the low side with
side inlet openings at frequent intervals. The drain connects with
a sump at the low point of the tunnel, from which a discharge pipe
is carried under the roadway of each tunnel to the New York River
shaft. Intercepting sumps with pumping equipment are provided
in all the river and land shafts.
The tunnel is lighted by electric lamps located in the side walls of
the tunnel immediately below the ceiling slabs. A continuous water
main is provided throughout the entire length of each tube, with hose
connections for fire protection and flushing at frequent intervals.
The walls are lined with white tile, care being taken to eliminate
all tile containing blue, green, or red tints, upon advice of a " color
psychologist," on account of its " depressing effects." The color of
the borders is a light orange. The ceiling is painted white.
The tunnel, with its twin tubes, 29 feet 6 inches in diameter, is the
largest subaqueous tunnel in America, exceeding by 6 feet 6 inches the
Pennsylvania Railroad tubes. On the New Jersey side, the diameter
of one of the tubes is increased to 30 feet 4 inches to meet ventilation
requirements. This exceeds by 4 inches the diameter of the Rother-
hithe Tunnel under the river Thames, London, England, which has
been the largest subaqueous tunnel in the world.
The shield method of construction was adopted for the Holland
Tunnel after careful consideration of other schemes, notably the
trench method. By the trench method, the work is conducted from
a plant floating in the river, and the tunnel is constructed either
under a protecting roof or floated into position and sunk in sections
in a dredged trench. The longest subaqueous tunnel built by this
method is the Detroit River tunnel of the Michigan Central Railroad.

It was recognized that in the excavation of a trench under the
Hudson River, there would be an unavoidable interference with a
great volume of river traffic. Fifteen hundred boats cross the line of
the tunnel daily. Such congested river conditions would make every
dredge or other machine working in the tunnel an obstruction to
traffic. Collisions would be frequent, increasing the time and cost of
the work, with danger both to shipping and to the equipment of con-
struction. Storms, fog, and ice would cause a discontinuance of sur-
face work for at least two months of each year. At the New York
end, a large mass of ledge rock, involving blasting and removal at
great depth, would be a serious obstacle to open-trench excavation
under water.

Since there was a real hazard involved in carrying on operations
from a plant anchored in midstream, the shield method was clearly-
called for. In addition, silt conditions in the Hudson River were
regarded as extremely favorable to this method. In a trench tunnel,
soft material greatly increases the volume of excavation, while in the
case of a shield tunnel this material is most easily excavated. If the
silt is not shoved aside by the shields, it is easily disposed of through
the tunnel. The shield may be closed with the exception of certain
openings through which the material is squeezed into the tunnel as
the shield advances.
The first contract provided for the sinking of two land shafts,
one at Washington and Canal Streets and the other at Washing-
ton and Spring Streets, New York City. They were sunk by the
compressed-air method.
The double steel walls of the caissons were filled with concrete as
the caissons were sunk. This added to their weight when sinking
weight was needed, and at the same time completed the structure of
the walls. In addition to this concrete, weight for sinking was ob-
tained by storing the excavated material from the working chamber
on the roof of the chamber as the caisson went down. This necessi-
tated handling the material a second time, but gave the desired
weight and permitted the lowering of the caisson without greatly
reducing the air pressure in the working chamber, thereby prevent-
ing loss of ground.
Upon the removal of the compressed air, the bottom seals of the
caissons proved to be water-tight. The shafts were now ready for
the building of the shields preparatory to the beginning of shield
tunneling. Temporary bulkheads were provided in the west side
walls to permit the passage of the shields, and in the east side walls
to connect with the approach section which was to be constructed
by excavation from the surface.
This work was followed by placing under contract the entire un-
der-riv'er portion of the tunnel. Power plants had to be constructed
to produce low-pressure air for caissons and tunnel, high-pressure
air for the operation of grouting machines, air drills, and hoisting
engines used below the surface, and hydraulic pressure for operat-
ing the jacks used in driving the shield and for operating the erector
arm for building the tunnel lining.
Overhead gantries and dumping platforms for the receipt and
disposal of materials and buildings for housing the workmen had
to be provided. Pipes, through which compressed air would be
supplied to the tunnel headings, had to be laid to the shafts. On
the New Jersey side this involved laying low-pressure lines as large
as 16 inches in diameter, high-pressure lines, hydraulic lines, water

lines, electric cables, and telephone cables. Every facility had to
be provided, even an independent telephone system connecting all
parts of the work with the public telephone system.
Canal Street Park was made available as a site for the air-com-
pressing plant and engineer's field office. Pier 35 and adjacent slips
were used for the storage of materials and for the disposal of ex-
cavated matter from the tunnel heading. Overhead gantries con-
necting the shafts with the pier permitted traffic to the water front
in connection with the tunnel to pass above the city streets.
The first shield was erected in the Canal Street shaft. On Oc-
tober 26, 1922, compressed air was introduced into the shield cham-
ber, and tunneling was begun. Each shield was 30 feet 2 inches
in outside diameter, 16 feet 4 inches long, and the upper half was
equipped with a hood projecting 2 feet 6 inches ahead of the shield
proper. Five vertical and three horizontal walls divided the shield
into 13 compartments, through which the ground in front was ex-
cavated. It was equipped with thirty 10-inch jacks, having a
combined thrust of 6,000 tons. A hydraulic erector was used
to build the tunnel segments into a complete ring. The weight
of the shield, with all equipment, was about 400 tons.
The tunnel lining is composed of rings 2 feet 6 inches wide,
consisting of 14 segments, each approximately 6 feet long, with
a key 1 foot long, bolted together. Inside the lining is an inner
lining of concrete 19 inches thick. As the shield advanced and
the lining was erected behind it, the space due to the difference in
the diameter of the shield and the rings forming the lining was
filled by forcing a grout of cement and sand in equal parts into the
void under high air pressure. For this purpose each segment
was provided with a grout hole fitted with a screw plug. The lining
was made water-tight by placing hemp grommets soaked in red
lead around the bolts, and by caulking lead wire into grooves be-
tween the segments.
Shield driving requires extreme care and exactitude to keep to
line and grade. The position of the shield fixes the location of the
tunnel, and no correction can be made afterward. It is absolutely
essential that the slightest deviation of the shield from its theoreti-
cally correct position be known at once, so that measures may be
taken to remedy the error during the next shove. The shield is
guided by the operation of the jacks distributed around its cir-
cumference, omitting the use of those jacks in the direction toward
which the shield is to move.
Every precaution was taken to provide for the safety of the work-
men in the compressed-air chambers. A high emergency gangway in
the upper part of the tunnel led from the shield to the locks, for
escape In case of a blowout. Safety screens were installed to trap the

inrushing water. Fire lines were installed in the compressed-air
chambers. Fire is a real danger in compressed-air work on account
of the increased amount of oxygen present. As an indication of the
fire hazard, a candle, if still glowing when extinguished, will again
burst into flame.
The starting of the shields out of the caissons at the New York
land shafts was difficult because of the large diameter of the shields
and the shallow cover overhead. The material at this point was
granular, consisting largely of fine sand, which if undisturbed, held
air fairly well. As the shields were under the city streets, it was
impossible to increase the cover overhead. To avoid blow-outs at the
face with the consequent inrush of water, it was necessary to regulate
the air pressure carefully and to protect the face during each suc-
cessive step in excavating.
As a preliminary step to shoving the shields out of the caissons,
the circular steel bulkheads in the caissons were burned out in front
of the shields. The work was done by removing the steel in hori-
zontal layers, each layer carefully protected as the steel was removed
to avoid exposing a great area of the face to air leakage, especially
when the air pressure sufficient to dry out the bottom would be heavy
enough to cause a blow-out at the top.
Removal of the steel bulkhead was started, with the steel above
intact and with air pressure sufficient to dry out the bottom. After
the lower third of the steel bulkhead had been removed, a wooden
bulkhead was built in front of the shield, and the space between this
bulkhead and the ground ahead was packed with clay. The air
pressure was then reduced until it balanced the water pressure at the
top of the shield, and work was begun at the top, removing the top
plates and proceeding downward.
As these plates were removed, breast boards packed front and back
with clay were inserted to cover the exposed excavation. This work
proceeded down to the point where the bottom plates had previously
been removed, while at the same time the air pressure was raised step
by step to balance the water pressure. The shield was then advanced
against the wooden bulkhead at the bottom, compressing the clay
which was removed as the shield advanced, with the jacks reacting
against the cast-iron tunnel lining temporarily erected in the shaft.
In order to prevent the leakage of air around the hood of the shield,
an annular pocket was excavated ahead of the hood the full length
of a shove, and this pocket was packed with clay. This served a
double purpose: First, the hood, as the shield advanced, cut into this
clay and made a thorough seal in front against air leakage; and
second, by exploring the full length of the shove, assurance was had
that the shield would not pick up and drag timbers in front of it,
leaving open channels behind them through which air could readily

escape. The necessity of taking this precaution is evident when it is
considered that at this point there were but 14 feet of cover above the
shield to the street surface, and only 8 feet from the top of the shield
to the under side of an old brick sewer, which would readily allow
the air to escape from the tunnel heading.
As the tail of the shield left the caisson, grouting was at once
started to fill the annular space which the shield left outside the
tunnel lining. Every effort was made to keep this space fully
grouted, even to the extent of stopping the shield in the middle of a
shove to keep the grout up with the shield.
The method just described was later modified so that in the bottom
quarter of the shield, instead of packing ahead with clay, a fixed
wooden bulkhead was built in the shield, and the shield was advanced
into the fine wet sand with this bulkhead in place. This compressed
the earth, driving out the water, so that the material was firm and
could be excavated during the shove over the top of the bulkhead, or
through small openings cut in the bulkhead itself. This prevented a
free run of wet material into the bottom which is the ordinary method
of tunneling under the river.
The grouting previously described was continued, and not only
prevented an abnormal escape of air at the tail of the shield, but
also prevented settlement of the streets and adjacent buildings.
The buildings at the corner of West and Spring Streets settled
slightly, but at no time were they in need of shoring, nor were the
occupants disturbed at any period of the tunnel work. This was
the situation also with the New York Central tracks under which
the Canal Street tunnel was driven. The grouting was carried on
so effectively that it filled some of the old sewers in the vicinity
which later had to be cleaned out.
The Canal Street shield passed very close to a cofferdam around
an excavation for a sewage treatment plant, and it was evident from
the first that great care must be exercised in driving the tunnel past
this location. At the nearest point the shield was within 5 feet of
the steel sheeting of the cofferdam, with the bottom of the sheeting
at about the springing line of tunnel. On November 30, when the
shield was about 40 feet away, it was noticed that sand and water
were being forced through the sheeting into the cofferdam by the
air pressure from the tunnel heading. In about 2 hours approxi-
mately 150 cubic yards of earth had been blown into the excavation
from behind the sheeting, and it was plain that not only was the
cofferdam in danger, but the continuation of tunneling operations
would be hazardous because the cavities left in the ground provided
open channels for the leakage of air, which might have resulted in
a tunnel blow-out. It was decided that tunneling operations should
be temporarily suspended, that the steel sheeting of the cofferdam

should be left in place permanently, and the concrete walls of the
permanent structure placed immediately, being increased in thick-
ness to enable them to withstand the pressure from tunneling
Preparatory to tunneling under the river bulkhead, clay and other
material to prevent the escape of the compressed air from the tunnel
were deposited in the slip between the piers and on the landward
side of the river bulkhead to fill such voids as might remain around
the tops of the piles supporting the timber platform of the bulkhead
construction. Not only were the voids around the piles filled, but
the soft mud in the slip was displaced by the heavier clay, a firmer
material and better adapted to resist air leakage.
In this section great care was taken in excavating ahead of the
hood to be sure that all piles within the area of the tunnel section
were cut off before coming in contact with the shields. This was
done to avoid pushing the piles through the ground and leaving back
of them an open channel for air to escape. These piles extended down
to the springing line of the tunnel excavation, and as many as 30 had
to be cut off at one time in advancing the shield the length of one
ring. In this manner both shields passed under the river bulkhead
without accident.
The tunnels then entered the Hudson Biver silt. The front of the
shield was completely bulkheaded. Some of the lower pockets in the
shield were opened to allow a part of the material to enter the tunnel
as the shield was advanced. The balance of the material in excava-
tion was displaced bodily. At once it was noticed that there was a
tendency of the tunnel lining to rise behind the shield. This rising
always accompanied the movement of the shield; whenever the shield
was stopped the rising ceased. The difficult feature at this point was
that the shield was so heavy that it settled while the cast-iron tunnel
lining behind the shield rose, so that the shield at all times was below
grade while the tunnel lining a short distance back was above grade.
The bulkheads in the shield were moved forward to reduce weight
by lessening the amount of muck in the shield. This aided somewhat
in keeping the shield from settling and then more material could be
taken in through the shield. This procedure lessened the pressure on
the tunnel behind and reduced its tendency to rise. As the contract
required that a second tunnel bulkhead should be constructed in this
vicinity, the south shield was stopped after passing through 218 feet
of silt and the bulkhead was built. This bulkhead, which is typical
of all the bulkheads, is a concrete wall 10 feet thick, equipped with
the usual muck, man and emergency locks, and adds temporarily con-
siderable weight to the tunnel.
With this additional weight, the rising of the tunnel was some-
what checked and after tunneling a distance of 121 feet farther

in the silt the shield entered at the bottom of the sand layer
which overlies the rock, and thereupon all rising of the completed
tunnel during shield driving ceased. In the north tunnel, which
was driven through the same material after the south tunnel was
built, a larger amount of material was taken in through the shield at
the start, and while there was some rising of this tunnel behind the
shield, it was very much less than in the south tunnel. In neither
tunnel was the movement sufficient to endanger the structure.
The excavation in the part-earth and part-rock section just east of
the New York river shaft caisson was carried on by driving a short
bottom heading in advance of the shield, in which was placed a con-
crete cradle with steel rails embedded in it upon which the shield
slid. After placing the cradle the rock was blasted out for one or
two advances of the shield and then the soft material on top was
carefully excavated and supported by poling and breast boards.
The New York river ventilating shaft caisson was sunk by the
compressed air method in the river near the New York pierhead line.
It was built on launching ways, then launched and drydocked. After
concrete had been placed in the pockets surrounding the working
chamber, additional steel was erected, carrying it to a height of 55
A platform supported on piles had been built on three sides of the
site (the south side being open ready to receive the caisson), and the
caisson was towed to its position on the work. The caisson at that
time weighed approximately 1,650 tons. Upon arrival, additional
steel was erected and concrete was placed in the walls, the caisson
sinking as the additional weight was placed. Care was taken to
keep the center of gravity as low as possible to maintain the
necessary stability. When it had reached a depth of 35 feet, the
cutting edge encountered the river bottom, into which it settled at
each low tide, and weight was added with sufficient rapidity to over-
come the tendency to float on the subsequent rising tide.
No excavation was carried on in the working chamber until the
cutting edge had penetrated about 9 feet into the mud, as the weight
of the caisson displaced the material up to this point. Compressed
air was then introduced into the working chamber and the usual
shaft mucking operations started. At a depth of 69 feet below mean
high water, rock was encountered. This was taken out in lifts about
6 feet deep and the caisson was lowered by successive drops until it
reached its final position.
The upper half of the outside of the caisson, or the part which is
exposed to open water, was covered with water-proofing, which in
turn was covered with an 18-inch layer of protection concrete. An
additional protection is afforded in the upper portion by a granite
facing where the shaft is exposed to tidal action.


[Illustration] FIGURE 2.-Longitudinal section through tunnel heading showing construction operations. Below : Rear of shield showing erection of iron and
mucking in process; view from rear of shield with bolting and grouting in process; exterior view of concrete bulkhead showing air locks

After the caisson was sealed to the rock and waterproof, the east
and west shield bulkheads in both the north and south tunnel cham-
bers were burned out and both shields were driven through the cais-
son. A timber and concrete cradle of sufficient strength to carry
the shield was erected in each chamber and the shield jacked across.
After the shields had progressed a sufficient distance west of the
river shaft to permit tunnel bulkheads, these were built in each tun-
nel and placed in operation. After this, tunneling operations were
carried on from the river shaft, releasing the tunnels between the
land and river shafts for the placing of concrete lining.
The caissons for the north and south land shafts on the New
Jersey side were assembled and sinking started in the fall of 1922.
After the caissons had passed through the cinder fill of the railroad
yard, a timber crib filled with riprap was encountered which made
excavation extremely difficult. The timbers had to be sawed or
chopped into short lengths and some of the rock broken up.
The distance between the tubes on the New Jersey side required
the sinking of two separate river ventilating shafts. This presented
a problem due to depth of the bedrock, 250 feet as compared with 70
feet on the New York side. It was considered that the silt which
overlies the bedrock would not afford a satisfactory support.
Accordingly, it was decided to support the shafts by means of
steel casings 24 inches in diameter, filled with reinforced concrete,
extending from the bottom of the shafts to ledge rock. They were
made in lengths of 20 feet, threaded at both ends for couplings.
Three lengths were connected and one end lowered into the silt. The
silt inside the pipe was then loosened by churning with a 2,000-pound
bit, and the mud arid water bailed out. Excavation was continued
in this manner to a depth of approximately 20 feet below the bottom
of the pipe. The material was firm enough to prevent caving into
the hold. Another section of pipe was then added and the entire
section driven into the hole previously excavated.
The north tunnel shield east and the south tunnel shield west were
built first and started out from their respective caissons. After the
south tunnel shield west had progressed a sufficient distance to erect
a tunnel bulkhead, the face of the shield was bulkheaded and the
roof was removed from the south caisson and the south tunnel shield
east was erected. As soon as this shield was ready, the roof was
replaced on the caisson and the shield was started eastward, so that
at the close of 1923 two shields were tunneling eastward, and one
The method followed in starting these shields out of the shafts was
similar to that already described for the New York shields, except
that here it was not so difficult as there was adequate cover overhead.
After the roof of the working chamber had been replaced, the

girders in the side of the caisson, through which the shield was to be
advanced, were burned out, after which the plates were removed from
the invert to the springing line. The lower pockets of the shield
were then bulkheaded and the space between the pockets and and the
exposed face was filled with clay. After this, the remaining plates
were removed, proceeding upward from the springing line. A semi-
circular annular ring was cleared for the hood and packed with clay
into which the hood was forced when the shield was advanced.
The material at the face consisted of timber and riprap down to
the springing line, similar to the material encountered in shaft
sinking, making excavation very difficult. The stones in the crib
varied from 1-man stones to those three-quarters of a yard in size.
The voids between the stones were filled with soft black mud, which
did not offer sufficient resistance to prevent the escape of air, neces-
sitating the mudding up of the entire face with clay. As the excava-
tion was carried forward, the escape of air through the heading of
the north tunnel at times taxed the full capacity of the power house,
40,000 cubic feet of free air per minute.
On June 10, 1923, a small blow occurred at the face of the shield
and it became necessary to drop the air pressure sufficiently to allow
the water to flow into the tunnel before the blow could be stopped.
The progress through the riprap was very slow, as extreme measures
had to be taken to avoid blow-outs. After the shield had passed
through the old timber and riprap crib, the river bulkhead was en-
countered which did not offer any unusual difficulties.
Before tunneling through similar material in the south tunnel east
5,500 bags of 1:1 Portland cement grout were ejected through the
east shield bulkhead of the south caisson and six pipes were sunk
from the surface east of the caisson through which 140 bags of
1:1 Portland cement grout were placed. This grout displaced much
of the soft mud and filled the voids in the riprap and greatly facili-
tated the driving of the shield so that very little air escaped through
this material after it had been consolidated by grouting.
After about 60 rings were erected in each tunnel, the shields were
stopped to build tunnel bulkheads and to install cages at the shafts
and then tunneling was resumed. Immediately east of the river
bulkhead soft mud, considerably lighter than Hudson Eiver silt, was
encountered in the upper part of the excavation. In this material
the tunnel began to rise directly behind the shield and also to move
To hold the shield and the tunnel to the proper grade, it was
necessary to take in a certain amount of material through the shield.
Accordingly, the shield was advanced with the top pockets bulk-
headed and a large percentage of the excavation was permitted to
enter the tunnel through openings in the lower part of the shield.

This material had to be entirely removed after each shove before the
erection of the cast-iron lining could proceed and slowed down prog-
ress. In addition it was desired to retain this material in the tunnel
directly behind the shield so as to increase the weight of the tunnel
and reduce the tendency to rise.
To meet this situation a different method of tunneling was adopted.
The work was stopped and a steel bulkhead semicircular in shape and
fitting into the lower part of the tunnel was built to trail about 10
feet behind the shield, and four pockets of the shield immediately
above the springing line were equipped with hydraulically operated
doors. When the shield advanced, these doors were opened varying
amounts, depending upon conditions, to allow the material to flow
through the shield into chutes which cropped the silt back of the
trailing bulkhead. This method of tunneling permitted both the
shield and the tunnel to be kept on grade.
River-shaft caissons were built, launched, floated into position, and
sunk, as on the New York side.
On October 22, 1924, shield driving was suspended in the north
tunnel from the New York side and a bottom heading or junction
drift was started to meet a corresponding drift from the New Jersey
heading. On October 29, the rock barrier remaining between these
headings was blasted away. After this all tunneling operations were
conducted from the New York side, as the junction was much nearer
the New York shaft. The south tunnel headings were joined on
December 7, 1924. Work on the New York side was suspended and
the New Jersey shield driven to meet the New York shield.
In July, 1924, the placing of the concrete lining forming the road-
way and air ducts was started on the New York side in the north and
south tunnels between the land and river shafts. The concrete invert
was first placed in both tunnels from the land shafts to the river
shafts. The remaining concrete was then poured in nine operations.
Five types of collapsible steel forms in 60-foot sections, afterward
increased to 75 feet, supported and moved by carriages resting on
previously placed concrete, were used.
The approach tunnels from the land shafts to the open approaches
at Dominick and Hudson Streets, New York City, and at Provost
Street, Jersey City, were built by the cut and cover method as usually
employed in subway construction.
A visitor to the Holland Tunnel in 1924 has written the following
graphic and interesting story of the shield method of construction.
The invitation to inspect the tunnel read, "Wear old clothes and
bring your galoshes."
Such was the admonition of our host on a warm September evening in 1924.
But knowing our host, we complied without ado other than a casual lifting of
the eyebrows. Ten o'clock that evening found four of us being piloted toward
Canal Street and the administration building of the Vehicular Tunnel.

Chief Engineer Holland himself greeted us, and began an introduction to
this vast engineering project with maps, diagrams, and more maps and dia-
grams, till red lines showing tunnels, and blue lines showing traffic lanes, and
green lines showing river bed swam before our gaze. We nodded very know-
ingly, mumbled pleasantly that exquisite shades had been chosen for the
various lines, and moved on to the doctor's office.
Here we were introduced to the necessary procedure before going into
compressed-air chambers. Ears, heart, and blood pressure were examined.
As we were found physically fit, we were passed on to the wardrobe, where
we were presented with an assortment of khaki cover-alls and left to our own
discretion as to choice.
The first twinge of squeamishness about cleanliness was quickly dispelled
by the romatic second thought that the very men who were performing this mir-
acle under the river had worn these self-same garments. Then followed a
scramble for the most bespattered on the theory that such muck was a mark
of courage in dashing into subaqueous passages. Size was completely disre-
garded. Never in all the stages of dressing up to set forth for adventure in
my childhood days had I enjoyed more of a thrill as so arrayed we followed
our guides to the tunnel entrance.
Once inside we were amazed to find what a simple form such a complex
sounding work could assume. The tubes are made of cast-iron segments bolted
together. Fourteen of these sections are required to make a complete ring.
Each section weighs tons and is held in place by huge bolts weighing 10
" Oh's " and "Ah's " were vented as we continued our way to see the actual
excavating. We passed groups of men sitting about talking, laughing, and
playing cards, awaiting their shifts. Work was never stopped 24 hours a day,
7 days a week. (With an investment of $42,000,000, it was imperative that no
time be lost.) We met car after car of excavated material on its way down the
temporary tracks to the entrance and out to be dumped.
At last we arrived at the great concrete bulkhead that sealed the compressed-
air section, separating it from the completed portion of the tunnel.
The bulkhead contained four air chambers or locks. Two large compartments
at the bottom of the bulkhead were equipped with tracks for bringing supplies
to the workers and for removing the excavated material. Two smaller cham-
bers were provided in the upper section for the workmen who on entering and
leaving the tunnel must be gradually brought from one pressure to another.
We entered one of these (only one was used normally, the other reserved
for emergencies) and saw the iron door clanged to and fastened. Then followed
lessons in equalizing the pressure inside and outside the head by holding the
nose and " snorting "-very much as one does when trying to expel water from
the nose after diving. The danger of the " caving in" of one's eardrums
was stressed, and we were warned to hold up our hand the moment the pressure
became too severe. This was the only way to attract the attention of the man
who turned on the compressed air, as the noise made even shouting inaudible.
We sat wilcl-eyed, expecting the hideous monster to leap upon us any minute.
The bark was worse than the bite. Twice we raised our hand and the pressure
was turned off until the pressure in our ears was relieved. When the 29-pound
mark was reached the door leading into the high-pressure section was opened,
and there we were in the very midst of the digging.
Once accustomed to the pressure, it was not noticeable, and we began a
siege of questions about the actual excavating.
This work was done under a shield, or movable head, slightly larger than
the external diameter of the tunnel. The shield was forced forward 2% feet

at a time, the width of a section, by means of 30 hydraulic jacks supported
against the end of the tunnel already built. Several of the jacks were then
removed and a segment was hoisted into place by a tremendous erector arm
till a complete ring had been added, and then the shield was forced ahead
again. Doors in the lower part of the shield allowed about 30 per cent of
the displaced compressed silt to enter the tunnel on each shove.
We stood watching the big burly men as they shoveled the dfibris into the
cars that carried it out through the lower air chambers. Not particularly
envious of them at such hard labor, we listened only half-heartedly to our
guide until he remarked that the automobiles we had seen parked at the
entrance belonged to these very " sand-hogs"; that they made high wages
and worked short hours. There are laws forbidding their working in com-
pressed air for more than two hours at a time for health reasons. Law like-
wise requires the company employing the men to furnish hot showers and hot
coffee for them when they come out.
From the digging we turned to watch the erector; two men tugging at a
mammoth wrench tightening the bolts; the grouting machine as it forced its
mixture with pressure beyond the segments to form a concrete shell for the
whole tube; and then to discuss the miracle that prevented the Hudson itself
from pouring in on us in one deluge. There we stood with only a few feet of
sand and gravel between us and the river.
"Chief" Holland and the rest of the engineers chatted with us as casually
as if it were a game of tiddle'-de-winks they were explaining, instead of an
achievement that even seeing denied believing. We picked up bits of rock for
souvenirs and continued gasping when one of our hosts turned questioner. He
asked if we could whistle.
Assuring him that whistling did not stump the modern girl, we inquired his
preference as to a tune. He consulted the other men, and after much deliber-
ation proposed to give us a big party on the condition that we whistle " Yankee
Doodle"-all five verses. With one accord lips were puckered and cheeks
distended. Our chagrin was only equaled by the laughter of our tormentors
as we puffed and blew in vain. The party was given for effort and not for the
results obtained against 29 pounds of pressure.
In quitting the compressed air it was necessary to put on fleece-lined coats
to prevent catching cold. We retraced our steps through the man lock, where
the pressure was reduced gradually back through the tube, and insisted on the
law requirement of hot coffee on signing off.
The problem of ventilation of the Holland Tunnel was unlike any
heretofore solved, both in character and magnitude. The only exist-
ing vehicular tunnels even approximately comparable to the Holland
Tunnel are the Blackball and Rotherhithe Tunnels under the Thames
at London.
The Blackwall, opened for traffic in 1897, has an under-river length
of 1,221 feet between shafts. It consists of a single tube 27 feet in
diameter with a roadway accommodating one line of traffic in each
direction and two sidewalks. Traffic counts in 1920 showed that the
maximum number of motor vehicles using the tunnel was less than
100 per hour.
The Rotherhithe is 30 feet in diameter, similar to the Blackwall
in traffic facilities, with an under-river length between shafts of 1,570


feet. Both of these tunnels are ventilated by the natural movement of
air through the shafts and portals. The Holland Tunnel, with a total
length of 9,250 feet, an under-river length of 5,480 feet, and a capacity
of 1,900 vehicles per hour in each direction, or 46,000 per day, obvi-
ously required something more than natural ventilation. To this
end the ventilation of the tunnel was studied under three heads:
1.The amount and composition of exhaust gases from motor vehicles.
2.The dilution necessary to render the exhaust gases harmless.
3.The method and equipment necessary for adequate ventilation.
The impurities in the atmosphere of a tunnel used by motor vehicles
are the product of the combustion of gasoline. If complete combus-
tion occurred, the carbon content in the gasoline would be in the
form of carbon dioxide, which can be tolerated in considerable quan-
tity without injurious effects. In a gasoline engine, however, com-
plete combustion seldom, if ever, takes place. The exhaust gases
contain varying amounts of carbon monoxide, depending on such
variable factors as the quality of the gasoline, conditions of car-
buretion, etc.
Carbon monoxide is a highly poisonous gas, injurious to health in
minute quantities if breathed for a long time, and if present in large
quantities is injurious even when breathed for a short time. Venti-
lation requirements are determined by the quantity of this gas in
exhaust gases. If sufficient fresh air is supplied to reduce this gas to
a safe percentage, other gases and impurities, such as carbon dioxide,
methane, and smoke, will also be diluted sufficiently. The first con-
sideration, therefore, was to determine the amount of carbon mon-
oxide that would be liberated in the tunnel.
Investigations were carried out at the Bureau of Mines experiment
station at Pittsburgh. The schedule called for the testing of pas-
senger cars and trucks of various makes and capacities. The tests
were made with cars loaded and light, standing with engine racing
and idling, accelerating from rest on level grade and on maximum
grade, running at 3, 6, 10, and 15 miles per hour on level and up and
down a grade of 3-1/2 per cent, corresponding to the maximum tunnel
grade. A total of 101 cars were tested. Gas samples were taken
directly from the exhaust pipe throughout the entire duration of
the test.
In general, the results showed that the exhaust gases contained
about 6.8 per cent carbon monoxide and 8.4 per cent carbon dioxide,
developing only 67 per cent of the heat value of the gasoline. About
one-third of the gasoline fuel was wasted through incomplete
Experiments to determine the proper dilution to render the
exhaust gases harmless were conducted at the Bureau of Mines
experiment station at Yale. They were performed in a gas-tight

chamber of 226 cubic feet capacity. Members of the staff spent
periods of one hour in air containing amounts of carbon monoxide
varying from 2 to 10 parts in 10,000. In addition, tests were per-
formed in a chamber of 12,000 cubic feet with an automobile en-
gine exhausting into the chamber. The duration of all tests was
one hour, whereas the length of time required to travel through the
tunnel at a speed of only 3 miles per hour is but 31 minutes.
The results of the test showed that when an automobile engine
is running properly the exhaust contains no substance that is in-
jurious to any appreciable extent except carbon monoxide. Gasoline
engines with cylinders missing, or when cold, over-supplied with oil
or gasosline, or smoking from any cause, may throw off disagreeable
vapors irritating to the eyes and nauseating to some persons.
The physiological effects of carbon monoxide are wholly due to
the union of this gas with the hemoglobin of the blood. To the
extent that hemoglobin is combined with carbon monoxide, it is by
that amount incapable of transporting oxygen to the body. This
combination of carbon monoxide with the hemoglobin is reversible,
so that when a person returns to fresh air the carbon monoxide
is gradually eliminated.
Of all physical signs and tests of carbon monoxide poisoning,
headache proved the most definite and reliable. Concentration of
gas too weak or periods of exposure too short to induce a headache
are to be considered harmless. No one had this symptom to an
appreciable degree after a period of one hour in the chamber with
four parts of carbon monoxide. With six parts the effect was usually
very slight, while with eight parts there was decided discomfort for
some hours.
Hence a uniform concentration of four parts carbon monoxide
in 10,000 of air is designed to afford not only complete safety, but
also comfort and freedom from disagreeable effects.
By the longitudinal method of ventilation, the entire tunnel would
be utilized as a duct for conveying air through the tunnel. Sufficient
air would be supplied through blower fans near one portal and would
enter the tunnel through a nozzle or nozzles at a velocity sufficient to
force it through its entire length.
If in a 29-foot tunnel the air were introduced into the north tube
near one portal through a nozzle having a cross-sectional area of 74
square feet, and were exhausted through the opposite portal, the air
would have a nozzle velocity of about 282 miles per hour. This
would produce a velocity of 72 miles per hour at points where the
roadway was occupied by a pleasure car and a truck abreast, or a
velocity of 51 miles per hour where there were no vehicles. Such air
velocities would be prohibitive in a vehicular tunnel, and the power
required to handle the air would be excessive.

In the distributive method of ventilation adopted for the Holland
Tunnel, the air is introduced into and exhausted from the tunnel
through a number of openings at frequent intervals leading from the
tunnel roadway. By this method fresh air is supplied at all points
throughout the tunnel. The air at any point can be controlled.
There is no discomfort or danger from high-velocity air currents.
The ventilation is not affected by traffic or the direction of the wind.
Exhaust gases are quickly diluted and removed.
The space above and below the tunnel roadway is ideally suitable
for air ducts. Fresh air, supplied by blower fans at the shafts, is
discharged from the main duct under the roadway through adjustable
openings into continuous expansion chambers on each side, thence

[Illustration] FIGURE 3. - Cross section - one tube of Holland Tunnel

through a continuous slot into the roadway. The air remains in the
tunnel an average of 1% minutes as it slowly ascends to the ceiling.
Exhaust fans located in the same buildings with the blower fans
draw the vitiated air through ports in the ceiling and thence through
the upper duct above the roadway, delivering it through stacks to
the outer atmosphere.
Experiments to determine the coefficient of friction for flow of air
in concrete ducts, to verify formulae used in computing the power
required for moving air through a duct from which air is taken off

at intervals, and to determine the power losses in bends or elbows
in concrete air ducts were conducted at the engineering experiment
station at the University of Illinois.
A concrete model, the linear dimensions of which were one-half
those of the lower duct of the tunnel, and 300 feet in length was used
for direct tests. Outlets with adjustable shutters to control the flow
of air were provided at uniform intervals on each side. Measure-
ments of air velocity and static pressure were made at three locations
in the duct, one 5 feet from each end and one midway. Tests were
run with all side ports closed and port pockets open at various inter-
vals, and with air velocities ranging from 1,000 feet to 6,000 feet per
minute. A total of 186 blowing tests and 17 exhausting tests were
run from which to determine the coefficient of friction.
On a full-size model of the expansion chamber proposed for the
tunnel, tests were made to determine the proper shape of the chamber
and the shape and size of the slot which would give a direction of air
flow high enough not to raise dust from the roadway and low enough
not to short circuit the fresh air to the inlets into the vitiated air duct
over the roadway. These experiments also gave the minimum static
pressure required to discharge the requisite quantities of air through
the slots at different locations in the tunnel. A total of 112 tests were
made on various shapes of expansion chambers and various widths of
slot under the several conditions to be met in the tunnel.
Experiments on elbows were made in two parts: On galvanized
iron single and compound elbows constructed to one-tenth the in-
terior dimensions of the elbows to be used in the tunnel, and on con-
crete compound elbows to one-half the interior dimensions of those
planned for the tunnel ducts.
To verify under tunnel conditions the amount of carbon monoxide
produced by automobiles and the physiological effect of exhaust
gases, an experimental tunnel was constructed in the workings of a
coal mine at Bruceton, Pa. It was located about 1,000 feet from the
entrance to the mine and about 135 feet from the surface. The
tunnel had a driveway 8 feet by 9 feet wide, with continuous air
ducts above the ceiling and below the roadway. It was oval in
plan, with a major axis of approximately 135 feet and a minor axis
of approximately 110 feet, giving a roadway length of 400 feet.
Air for the test was supplied by the mine fan, belt-connected to a
steam engine and operated outside the mines. The fan operated
normally exhausting, giving upward ventilation in the tunnel.
Downward ventilation was accomplished by reversing the direction
of the air currents through the reversible housing of the fan, which
then operated as a blower.
In the upward ventilation system, air entered the duct under the
roadway, passed through adjustable port openings into the con-

tinuous expansion chambers on either side of the roadway, thence
into the driveway. In the downward system, air was delivered to
the duct in the ceiling, thence through the ports into the upper ex-
pansion chambers from which it entered the roadway.
A total of 17 tests were run with cars varying in number from
1 to 8, with concentrations of carbon monoxide in the driveway from
0.5 to 9.4 in 10,000 parts of air, at various temperatures and humid-
ities, and various methods of transverse ventilation. The tests veri-
fied the earlier conclusions, and demonstrated that with upward
ventilation the exhaust gases crossed the breathing plane of per-
sons in the tunnel but once, while with downward ventilation they
crossed this plane twice. There was also a lower concentration of
carbon monoxide with upward than with downward ventilation.
Valuable and necessary as were the experiments required to de-
termine the various factors involved in the problem of adequate
ventilation for the Holland Tunnel, the data resulting from these
preliminary investigations had to be crystallized into tangible units
of ventilating equipment.
These are the 84 giant Sturtevant Silentvane fans which are the
very lungs of the tunnel. Without such fans blowing in fresh air
and exhausting the vitiated air the tunnel could not be made to
The Sturtevant Silentvane fans are installed in the ventilation
buildings, of which there are two on each side of the river, one at
the pierhead line and the other inland. Each land shaft ventilates
four sections of tunnel, the adjoining portal sections of each tube,
the whole intermediate section to the pierhead shaft where traffic is
on a downgrade, and one-half of the parallel section where it is on
an upgrade. The buildings over these shafts contain four indepen-
dent sets of blower and exhaust fans. The pierhead shafts ventilate
three sections of tunnel, one-half of each of the 3,400-foot river sec-
tions and one-half of the intermediate section where traffic is on the
upgrade. In all there are 14 sets of blowers and 14 sets of exhaust
fans. Dividing the upgrade sections of the tunnels into three parts
gives added ventilation where the greatest amount of carbon monox-
ide is expected.
There are 28 ducts-14 blower and 14 exhaust, connecting the va-
rious sections of the tunnels with the ventilating buildings. Each
duct is equipped with three fans, two of which, when operated to-
gether, will supply the maximum quantity of air required. Their
capacities range from 81,000 to 227,000 cubic feet per minute and
they operate at static pressures varying from 0.6 to 3.75 inches of
water. This range in pressure and capacity is due to the great
difference in length of tunnel ventilated by different sets, those at
the outside of the pierhead shafts having 1,700 feet to serve while

the inside fans have only 700 or 800 feet. These fans, during an
hour of heavy traffic, will handle 84,000 tons of air, or 1,400 tons per
minute. They provide for changing the air in the tunnel 42 times
per hour.
The fans are of the backward curved-blade type. Under differ-
ent conditions, one, two, or three fans may be operated on one tunnel
Duct at any one time.

[illustration] Figure 4. Typical vertical section through ventilating building

They are electrically driven by wound-rotor motors with resist-
ance in the circuit to make it possible to run them at variable speeds.
The combined capacities of the motors is approximately 6,000 horse-
power, two-thirds of which will be in operation at times of maximum
load and one-third in reserve. Chain drives are to be used to make
possible speed adjustments or changes in the motors as well as on
account of the space limitations in the ventilating buildings.

The placing of the fans is varied to suit the local conditions in the
individual buildings. Generally, the exhaust ducts are at the corners
of the buildings and supply ducts are in the central portion. Con-
sequently the compartments containing the exhaust fans are located
near the corners under the exhaust stacks, leaving the central por-
tions of the fan floors free for intake fans, and the central section of
each outer wall for the air intakes. The intakes are made sufficiently
large to give low velocities through the louvres.
The louvre blades are made of heavy wire glass to give light to the
interior of the buildings as they take up most of the space otherwise
available for windows. Heavy bronze screens protect them and also
serve to keep out birds.
The arrangement whereby fresh air is drawn in through louvres
high up on the sides of the buildings and exhaust air is forced out
through stacks which extend 20 feet above the roof insures a complete
separation of fresh and vitiated air.
The intake fans and their motors are situated in the open portions
of the fan floors where they are accessible. The exhaust fans are, of
necessity, inside of chambers at the top of the ducts. Their motors,
however, are out on the main floor, the drive shafts being run in to
the fans through close-fitting collars in the side plates of the duct.
Access to the fans is provided through air locks equipped with air-
tight doors which can be opened against the unequal pressure by
wedge latches which force the doors open sufficiently to break the
Each duct is Equipped with a damper which may be closed when
the fan is shut down so that air from the other fans will not be short-
circuited through the idle fan. These dampers are motor operated
from the control room and are equipped with limit switches.
An unusually flexible system of power supply has been worked
out based on the facts that all the motors are in groups of three, also
that the maximum power equipments are less than the capacity of
the minimum size power cables installed by the local companies.
Three cables from the New York side and three from the New Jersey
side are run to the bus bars in each ventilating building, thus giving
one motor in each set a separate cable connection to power supply on
each side of the river. Interconnection at the bus bars makes it pos-
sible to cut in any or all motors on each cable. Thus connected, each
motor may be supplied with power by six independent cables, each
capable of carrying the entire tunnel load; and, as there are at least
two independent sources of power at each end of the tunnel, con-
tinuity of power supply is absolutely assured.
As the transformers are located in the ventilating buildings where
smoke from an oil fire might be drawn into the ventilating system,
air-cooled instead of oil-cooled transformers are used.

Each fan is provided with a control switch at the motor for
emergency or repair use. Further local control is provided at the
switchboard in each ventilating building, and complete operating
control is provided at the main switchboard in the administration
building where, by a system of signal lights, it will be possible, at
all times, to tell what motors are in operation.
Air from the intake fans is forced down into the longitudinal
duct under the roadway of the tunnel. From there it is fed through
flues 10 to 15 feet apart into a continuous expansion chamber above
the curb line at each side of the roadway, the flow of air into this
chamber being controlled by adjustable slides over the flue openings.
The outer side of the expansion chamber is a copper-steel plate
which can be adjusted to give an opening of widths varying from
% inches to 1% inches through which fresh air flows into the
Vitiated air is drawn off through openings through the ceiling
into the exhaust ducts. These openings are spaced 10 to 15 feet
apart and are from 3 to 6 feet long. They, also, are provided with
slides by which the opening can be adjusted to meet the local re-
quirements for air circulation.
By this arrangement of supply and exhaust ports, fresh air sup-
plied to the roadway mixes with the warmer gases and rises to the
ceiling where the exhaust ports are located.
There will be no longitudinal movement of air in the tunnels
except that induced by the movement of vehicles, nor will there
be any objectionable winds such as would be created by longitudinal
ventilation. Tests made with smoke bombs showed that even large
quantities of smoke will not spread far from the point of origin,
but will rise quickly to the ceiling and be taken out. Similiarly,
in case of a fire the hot gasses will rise to the ceiling, where they
will be drawn off. There will not be the same danger of spreading
the fire from car to car as there would be with longitudinal
As part of the studies for the ventilating equipment, numerous
tests in relation to fire were made, both in the test tunnel at Bruce-
ton and at the laboratories of manufacturers of fire-fighting equip-
ment. These tests included the burning of an automobile drenched
with gasoline and with gasoline spilling from a hole in the tank
on the car to determine how quickly such a fire could be put out
with the hand extinguishers to be placed in the tunnel.
As a check upon the air conditions in the tunnel, automatic carbon
monoxide recording devices are installed in each exhaust duct which
will make a continuous analysis of the gases and record it graphically
in the control room of the administration building in New York.
There, by observing the chart, the operator can increase or decrease
the fresh-air supply as traffic conditions change in the tunnel.

[Illustration - map] FIGURE 5. - Location plan of Holland Vehicular Tunnel

That the construction of the Holland Tunnel was no easy task is
evidenced by the great increase in both time and money required for
its completion. The original plans called for an expenditure of ap-
proximately $28,000,000 and for completion in 1924, or three and one-
half years. Actual expenditures have run 50 per cent greater, and as
this is written, the opening will not be until the fall of 1927.
Yet this is not surprising. Although the shield method of con-
struction has been described in this story as if it were a relatively
simple operation, many difficulties had to be overcome in bringing
the work to a successful conclusion. The proceedings involved in the
taking of real property at entrances and exits, changes in the grades
of streets, the closing of a portion of Eleventh Street in Jersey City,
negotiations with the railroads at the Jersey City end for the acquisi-
tion of parts of the railroad yards, all took time. It was not always
easy to harmonize the views of the State Commissions. Alterations
necessarily had to be made in the preliminary plans as further infor-
mation resulted from investigation and experience.
That the undertaking cost the lives of its first two chief engineers,
not from accident, but from the drain on their vital energy, is perhaps
the most striking evidence of the magnitude of the undertaking.
[end text]

Genius of the Holland Tunnel and its first chief engineer, in memory of whom the tunnel was named.
Plate 2. Site of the Holland Tunnel, Looking West from New York City.
Plate 3. Sectional View of Holland Tunnel Under the Hudson River, Looking Toward New York City.
Plate 4. Milton H. Freeman.
Who Succeeded Mr. Holland as Chief Engineer, and who died in 1925.
Plate 5. Holland Tunnel and Hudson & Manhattan Railroad Tunnel
Full-sized section of Holland Tunnel (diameter 29 feet 6 inches) and full-sized section of Hudson & Manhattan Railroad Tunnel (diameter 16 feet 7 inches). Rings weigh 16,630 pounds and 5,670 pounds per linear foot, respectively.
Plate 6. Assembling Shield in Canal Street Shaft
View looking down into shaft, showing bulkhead in west side wall.
Plate 7. Land Shaft Cassion at Spring Street, New York City
Showing steel bulkhead in west side wall through which shield advanced after erection.
Plate 8. Shield, south Tunnel, Canal street at West street. New York City
View of rear end of shield in place and temporary bulkhead. Tunneling operations temporarily suspended and air pressure removed in order to remove shaft deck and
place cages in shaft and air locks in tunnel.
Plate 9. 1. Gang of "Sand Hogs" in line
Waiting to check in for work in compressed air. Canal Street land shaft and air locks, South Tunnel.
2. Hauling a Car of Muck out of a Muck Lock, South Tunnel
Plate 10. 1. Concrete Roadway
Beginning of sidewalk, and reinforcing of sidewalk, North Tunnel. View shows construction track on roadway and roof rebolting and calking platform.
2. Concrete Bulkhead and Locks
South Tunnel, Canal Street, New York City.
Plate 11. New Jersey River Shaft caisson, Just after Leaving the Ways
Caisson launched at Mariners Harbor, Long Island, floated into position and sunk.
Plate 12. Tightening Bolts in Tunnel Lining, North Tunnel, by Means of Ratchet Wrench. Each Bolt Weighs 10 Pounds
Plate 13. Erector Arm
Swinging iron segment into place in tunnel lining, South Tunnel New York City.
Plate 14. Holing Through
Tunnel superintendent Harry Redwood, of New York side, shaking hands with Norman Redwood of New Jersey side, North Tunnel.
Plate 15. Curve in South Tunnel under West Street, New York City (Radius 1,000 Feet). Showing Completed rings of Cast-Iron Lining
Plate 16. Concrete Construction in south Tunnel
Showing a typical cross section of concrete lining and details. The upper and lower arcs of the tunnel form the ventilating ducts.
Plate 17. Model of the Holland Tunnel Showing Many of the Hidden Details
[text of labels keyed to elements on the model]
Plate 18. Rest and Refreshment in Rotherhithe Tunnel. River Thames, London, England
Plate 19. Approach to the Blackwall Tunnel, River Thames, London, England
Plate 20. 1. East Blower Air Duct
In land ventilation building, New York City, showing curved back and vanes.
2. East Blower Air Duct
Land ventilation building. New York City.
Plate 21. 1. Fresh-air Duct in south Tunnel, New Jersey Side
Showing the beginning of the transition from its position under the roadway to its position alongside the tunnel.
2. Tile and Bronze Work
Left to right: Bronze door to relay niche with telephone and fire-alarm boxes on each side; tiled refuge niche with fresh-air outlet on each side, two fire-extinguisher niches; tiled opening to mid-river sump.
Plate 22. Land Ventilation Building
West side of Washington Street, Canal Street to Spring Street, New York City.
Plate 23. River Ventilation building
Pierhead line between Piers 34 and 35, North River, New York City.
Plate 24. Special Apparatus Erected at Hyde Park Plant of B.F. Sturtevant Co.
Used in testing Sturtevant silent-vane fans for the Holland Tunnel.
Plate 25. Sturtevant Silent-vane Fan Wheels for the Holland Tunnel
Plate 26. One of the 84 Sturtevant Silent-vane Fans which are the Lungs of the Holland Tunnel
Plate 27. Exhaust Fan Unit in New jersey Land Ventilation Building
Showing typical arrangement of exhaust fan, motor, chain casing, resistors, control cabinet, and local control box.
Plate 28. Condensation in North Tunnel
View showing dry condition of roadway east of air duct bulkhead where fans were in operation and wet condition of roadway west of air duct bulkhead where fans were not in operation.
Under whose direction the Holland Tunnel was brought to successful completion.
Plate 30. 1. Model of Entrance to Tunnel, New York City
Looking north-northwest across entrance plaza which comprises north half of block between Broome and Watts Streets.
2. Model of Exit from Tunnel, New York City
Looking northwest along Canal Street.
[end plates]

People Holland, Clifford Milburn
Freeman, Milton H.
Singstad, Ole
Redwood, Norman
Redwood, Harry
Date 1930-1931
Year Range from 1931
Year Range to 1931
Search Terms Holland Tunnel
Hudson & Manhattan Railroad
Caption 01 pg 577: title - The Eighth Wonder: The Holland Tunnel Vehicular Tunnel
Imagefile 085\20110050064.TIF
Classification Engineering
Government & Politics
Historic Sites