• Amanda Russell

Building the Stairs: Construction in Ha'ikū Valley

Updated: Jul 10, 2019

"One of the fabulous tales of the war now being made public concerns the great navy radio-sending station built in the utmost secrecy in Haiku valley, windward Oahu, starting in 1942..."

Honolulu Star-Bulletin, oct 25, 1946

Construction of the US Naval Radio Station in Haʻikū Valley was one of the most complex and perplexing jobs of the Pacific offensive. The facility was classified top secret and there was no discussion with the Army or the operating committee of the Navy. There was no real model to follow for engineering construction, and the terrain was extremely rugged and often dangerous to work in.

Along with the shortage of workers and materials there was also an atmosphere of heightened stress in Hawaii following the trauma of the attack on Pearl Harbor on December 7, 1941. The community was struggling with the tension of living under martial law, including black outs, and curfews. Barbed-wire was strung along the beaches as a vivid reminder of how vulnerable the islands were to attack.

Early in 1942 a proposal to build a pioneering radio communications system was put forward. The proposal was headed by Commander Hord, Radio Materials Officer at CINCPAC; BuDocks (Bureau of Yards and Docks) engineers, Lt. Commander R.M. Belt, Lt. Butzine, and Lt. Thatcher, engineers of the Radio Corporation of America, including Mr. McKesson; and the New York designing firm of Gibbs and Hill (Woodbury, 1946:349). The project was a result of extensive combined efforts.

Drawings were prepared by the District Public Works Design Section, and by the contractors’ design department. Mountain-top anchorage, method of raising and lowering the aerial systems, and layout of roads and trails were designed by the contractors. The aerial system, the antenna ground system, the transmission lines for carrying radio-frequency power from transmitter to antenna and ground system, and the transmitter layout (including suggestions for main and auxiliary building layouts) were designed by R.C.A. communications, Inc., in connection with their order to furnish radio-installation material and field-engineer supervision on installation. R.M. Towill, of Honolulu, was employed to make preliminary site surveys (U.S. Dept. Navy, 1944: A-816).

Certain portions of the work were later completed by construction battalion forces including, “installation of the 600-kw diesel standby generator in the bombproof transmitter building, installation of several poles on the transmitter building on the Pali summit, and installation of a short length of underground cable” (U.S. Dept. Navy, A-829; 1944).

The construction was difficult due not only to the sharp terrain but also due to extra pressures created by the war situation. Materials and labor were difficult to come by. “Before the Japanese attack, all barracks were constructed of reinforced concrete, but after the outbreak of war all personnel facilities were built of wood, to conserve critical materials” (U.S. Dept. Navy, 1946:139).

This was clearly the case at the Naval Radio Station at Haʻikū Valley. The barracks, mess hall, and other basic structures were of frame construction. Generally the most important buildings, those vital to the successful transmission of long range messages, were constructed of reinforced concrete. These were the bomb-proof transmitter building, helix building, intra-island communication transmission building, and the commercial electric supply building. On the island tensions rose as the need for rapid construction increased and supplies were limited. Often when supplies did arrive they were in poor condition and further time was needed to repair them.

The choice of Haʻikū Valley was based on its topography. In order to build the facility a site was needed that had two mountain walls rising as vertically as possible, both a similar height over 2,000 feet, and with flat land in between. By studying a Coast and Geodetic map it was determined that the walls of Haiku Valley eroded into a series of cliffs like in an amphitheater open towards the ocean, would be ideal. The valley was covered with dense vegetation and large lava rocks. Huge twisted hau trees were tightly knit throughout the valley. Axes were used to chop through the hau trees as the first expeditions were made though the valley (Woodbury, 1946:350).

Beyond clearing the valley, the largest problem facing the construction of the radio station was finding a way to send men to the top of the cliffs. The cliffs that semi-encircle the valley range between 1,800 and 2,850 feet. In many places the cliffs rise vertically at a 90 degree angle [sic]. To further add to the problems of scaling was the fact that the dirt was often either crumbly lava or was muddy and unstable from the high amount of rain and fog.

Image courtesy of the US Geological Survey
Image courtesy of the US Geological Survey

The contractors had considered sending back to the mainland for professional high-scalers when it was brought to their attention that two such high-scalers were working on another tough military project in nearby Red Hill. The two men, Bill Adams and Louis Otto, under the leadership of rigger Ray Cotherman, were known for their fierce determination and courage in conquering the heights. They began their climb up the steep slopes with one coil of rope, a rock pick-sledge hammer, and some three foot steel pins (Woodbury, 1946:356).

​The story of their ascent and the construction that they opened the door to has been documented in David O. Woodbury’s book Builders for Battle and magazines such as Reader’s Digest due to the harrowing nature of the work (Honolulu Star Bulletin, October 7, 1950).

Bill Adams and Louis Otto. Image courtesy of David Jessup
Bill Adams and Louis Otto. Image courtesy of David Jessup

The climbers began up the south, or Pali, side of the valley. The two men worked their way up the steep rock by driving in one spike, standing on it and driving in another, attaching the rope to that spike and then pulling themselves up to drive in another spike. When the spikes were used up for the day they returned to the bottom and began again the next day. It took Adams and Otto 21 days before they reached the top. What the two found was that the top of the cliffs were, in effect, a ‘razor top’. Nowhere was the top more than 12 feet wide and in most areas it was 4 to 5 feet wide (Woodbury, 1946:356).


During this time plans were being made for the construction and design of the facility. On May 14, 1942, a summary of cost was written by the five members of CPNAB for an estimate of a radio transmitting station. The estimate contained a detailed listing for the transmitter building which included the cost for clearing the site, 850 cubic yards of concrete and 136,550 pounds of reinforced steel to construct the bomb-proof structure. It also included the cost of installing a 10 ton crane (CPNAB, Estimate, 1942). This crane currently remains in place at the transmitter building at the Omega Station at Haʻikū Valley.

A naval engineer attends to one of the antennae housed in the Omega Station in Ha'iku Valley.
A naval engineer attends to one of the antennae housed in the Omega Station in Ha'iku Valley.

The estimate for the design and construction of the antenna’s anchors included costs for cableways to the ridges, lateral transport along the ridges, hoists and hoist houses, rigging gear, special tools, excavation and rock trimming, and a large sum for contingencies. This initial estimate attempted to prepare for the ability to make full use of the radio station as soon as possible. It allowed for not only the construction of the transmitter building and cableways but also telephone installation, roads, a fence, topographic survey, and quarters for 35 men. The equipment would be shipped on the mainland by railroad freight to California, by ocean freight to Hawaii, by truck from Honolulu to the site, placed in storage at the site, and it was planned that guards would be hired to protect the materials. The total estimate for the transmitter building was $598,300 and included 10.24% for accident insurance and taxes, 20% for the job overhead, 5% for the operating base office in Alameda, and a 5% fee for the contractors (CPNAB, Estimate,1942).

 On May 25, 1942, the valley was sketched out by Lt. Butzine (Butzine, 1942). The sketch showed the land area that the Navy hoped to lease for the station in order to ‘stretch aerial wires’. After some study, on June 11, 1942, BuDocks authorized acquisition of the land instead of leasing the area, with a decision that the land would be leased temporarily in order that construction on the radio station could begin as soon as possible (PACDIV:9). It was not uncommon for months to go by before legal acquisitions were made during the time of the outbreak of the war.

Lt. Butzine also noted on his 1942 rendering of Haʻikū Valley and its proposed aerials: "There has been no decision to date of the final locations of Aerials. BuDocks has recommended an overall spread of 1700 feet but the Officer in Charge has recommended that the maximum spread not exceed 1200 feet to simplify construction. It is essential to construct small helix houses under the aerials as well as installing certain ground and telephone poles" (Butzine, 1942).


This early sketch indicated both options of placements of aerials on the ridges, 1200 and 1700 feet apart. It also placed four helix houses near the center of the valley, one for each of the cables, a transmitter building at the east of the helix houses near the entrance to the valley, and a building that would serve as quarters for 35 men, also near the entrance and the transmitter (Butzine, 1942).

Over the next months the Contractors, PNAB regularly corresponded with the R.C.A. engineers in New York. Due to the fact that the project was experimental all aspects of the construction were up for revision as the combined efforts of the workers resulted in innovative solutions to handle each new problem as it arose. On June 23, 1942, the Plant Design Superintendent, J.L. Finch from R.C.A. Communications, Inc. wrote to the CPNAB regarding further specifics of the anchor and aerial placement. At this point R.C.A. was basing its information on photographs and maps they had been sent on May 19, 1942, and a field representative they had in Hawaii.

The serial system is to consist of four cables roughly parallel and strung between mountain tops in a northwesterly-southeasterly direction and with down leads hung from a point near the center of each. The northwest end anchorages are to be located on a ridge shown on enclosure C having an elevation of slightly over 2500 feet and with a length of about 630 feet in a northeasterly southwesterly direction. The Southeast end anchorages are to be located with two of them straddling the summit named “Puu Keahiakahoe” with two then to have an elevation of slightly above 2750 feet and the other two on a step along the ridge in a northeasterly direction from the summit and shown to have an elevation slightly above 2450 feet.


Numbering these cables from the northeast, according to the present plan the northwest end anchorages will be spaced as follow:-between No.1 and No.2 100 feet, between No.2 and No. 3 400 feet and between No.3 and No.4 100 feet. Similarly at the southeast end the anchorages will be spaced as follows:-between No.1 and No.2 100 feet, between No.2 and No.3 700 feet and between No.3 and No.4 100 feet (Finch 1942:1).

These R.C.A. measurements would make the aerial span only 900 feet at its widest, well within the recommended 1200 foot length. In the same correspondence there were specific discussions of the type of steel wires to be used and the positioning of the insulators, downleads, and counterweight placements (Finch 1942:2) The general idea of the plan was followed although the specifics changed during the construction.

 On August 4, 1942, R.C.A. wrote the Bureau of Ships a letter regarding the measurements of the transmitter room. They recommended a “possible decrease in the height of the transmitter room to 21 feet and in the height of the coil room to 15 feet (R.C.A., 1942:1). It appears from this correspondence that it was determined that the helix room and the transmitter room would be in the same building instead of the prior proposal of four helix buildings that would have been under each antenna cable. By October, 1942, the correspondence between Finch and CPNAB reflects a progression in the construction as it was more concerned with details, such as the specifics for duct trenches and trench cover details.


While engineering ideas progressed many workers were continuing to solve the problem of access to the cliffs, necessary to build the anchors. First, simple sections of ladders were brought up along the south side of the valley and attached to spikes in the mountain. The workers called this a “sissiesâ climb” and with it the first group of men could gain access to the ridge and search for the proper placement of the aerials and research building conditions (Woodbury 1946:356).

Access was found to the North Ridge by the use of a local pathway that approached the rim from the opposite side of the valley, the Red Hill approach. Apparently the pathway had been built earlier by the Civilian Conservation Corps, a group organized by the New Deal government to create jobs for young men. It was used both by hikers and bikers up till the time of construction of the radio station (Honolulu Star Bulletin, Oct. 7, 1950). Due to the highly classified status of the project the Navy had not notified the Army, the territorial board of agriculture and forestry, or any other relevant parties. Upon discovering that civilians were using the trail and observing the construction, action was taken to limit use of the trail strictly for Naval construction purposes (Honolulu Star Bulletin, Oct. 7,1950). Use of the back trail meant that there was no immediate need for scaling the cliffs and construction of stairs on the north valley cliffs, but a round trip using the back trail still required 23 hours.

Though delayed by week-long bouts of fog, engineers studied the rim and obtained distance and altitude information for the aerials. Geologists studied the mountain and were relieved to find that the muddy exterior and crumbly base were covering a solid foundation for the anchors. This was of particular concern because the five proposed anchors would be supporting a pull of 25,000 pounds each. A letter dated May 19, 1942 stated:

“The first 8 inches were black humus; the next 3 feet were a red crumbly volcanic conglomerate, and the last foot was grey rock and earth partially decomposed. It is believed that the ridge formations are composed of hard lava rock and conglomerate masses under a covering of humus and disintegrated material. It is believed that anchorages of expansion bolts bored into the rock, such as used in hard granite walls, would hold, due to the broken texture of the rock . . . “(U.S. Dept. Navy, 1944: A-818).

It was determined that they would not use the standard procedure used in stone of expansion bolts but instead would use concrete anchors (Woodbury, 1946:356).

While the exploration continued on the ridge, efforts were ongoing to improve the climb up the south wall. The ladders were replaced with wood steps and catwalks. The ladders were constructed in 6 foot sections using 1″x6″ sections of steps fitted into carved sections of long side boards to form a flush ladder section. Eventually there was 8,050 feet of stairway. 3,500 feet of wood ladders were built by the CPNAB in 1943 along the southern access of Haiku Valley. 950 feet of galvanized steel ladder was built along the ridge in 1947 by the contractor Walker Moodey. 3600 feet of glavinized steel ladder was built in 1952 along the northern access by contractor A.W. Yee (U.S. Dept. Navy, 1955). Records indicate that this northern portion of stariway was originally slated to be built as a wood stairway in 1944 (Klepper, 1944:1).

The stairs were originally called the Haiku Ladder. Pictured here are remains of the original wooden steps. Also in this image is John Goody, board member emeritus, during routine maintenance work on an authorized work day.
The stairs were originally called the Haiku Ladder. Pictured here are remains of the original wooden steps. Also in this image is John Goody, board member emeritus, during routine maintenance work on an authorized work day.

When the first section of stairway was completed on the southern access, it took approximately 3-1/2 hours to go up to the ridge using the stairs. Immediately following completion of the southern access work began to build cable cars and a tram system so that men, and most importantly heavy equipment, could be taken to the top of the ridge (Woodbury, 1946:357).

Engineers concluded that a small cable car, several men, and some materials would have an actual pull of nineteen tons. This, along with a determined safety factor meant that it must be built to withstand a fifty ton pull. This required 38,000 feet of inch wide steel wire rope. This amount of wire weighed twenty tons and left the obvious problem of getting it to the top of the mountain (Woodbury, 1946:357).

 Ideas for getting the wire to the top of the North Ridge were considered, including having it dropped by an Army bomber, but the final decision was to carry up a lighter wire loop. To do so, 52 men carried a 10,000 foot long, quarter-inch cable over 6 miles up the steep terrain by coiling up 100 foot sections, placing one over the shoulder of one worker, having him walk up 100 feet, and then loading up a second man with 100 feet of wire and sending him up, until finally the entire length was laid along the path of the mountain. The men reached the top by evening and camped at the ridge. The next day, one of the original mountain scalers, Bill Adams, volunteered to make the loop by taking the line down the north face of the cliff, where there were no steps or ladders. He spent the day sliding down and axing his way through the brush until he reached the bottom (Woodbury, 1946:359).

The South Ridge was easy in comparison as workers were sent up the wooden stairway carrying the quarter inch wire. At the top a small loop was made with a piece of boards placed in it for a seat and a man was lowered down in it to the bottom. With the light wires in place it was possible to use a diesel powered hoist to pull up the heavier steel wire (Woodbury, 1946:360).

 The cable car was supposed to carry either 500 pounds or three men at one time. Often the rule was overlooked.

” . . . at quitting time on top the men could not wait for rules. Everyone who could possibly hang on anywhere would get aboard and swing out over the chasm, singing and tossing around their rigger’s language.” The engineers did have one strict safety feature . . . the use of a ‘skookum block’ for each headend pulley. This block had a completely enclosed body and graphite-lubricated bearings, which made it impossible for the cable to jam or pull loose” (Woodbury, 1946:361).

Prior to raising the heavy wire the engineers wanted to test out the antenna. Temporary anchors were erected and a medium weight heavy cable hoisted. A small transmitter was set up in the valley and connected to the downlead (Woodbury, 1946:360).

The results of this test, along with the measurements and calculations of the Haiku antenna, came out on November 11, 1942. The test antenna had a 4,575 foot flat top and 1250 foot downlead. The final antenna was proposed to have a 4600 foot flat top, a 1453 foot downlead, and of course a stronger steel wire. Various measurements were taken such as ground resistance, comparisons with other low frequency stations, effective heights, measured capacity, and antenna efficiency (U.S. Dept. Navy, 1942b).

 The results of this test were a matter of great interest because the nature of the project was one of experimentation. This was clear in the introduction to the report:

"The erection of a low frequency antenna between mountain tops is a departure from normal antenna design. Previous experience with such antennas has been extremely limited, and because of this, it was impossible to estimate accurately the performance of such an antenna. The most important unknown was the radiation resistance value upon which the efficiency of the antenna depends" (U.S. Dept. Navy, 1942b).


The conclusion of the report indicates that the results were not as notable as was expected but did rate high enough to serve its intended purpose.

These tests and calculations show that the Haiku antenna as designed will be a good low frequency antenna. The test shows that the performance will be considerably below the expected performance based on preliminary calculations. However, it will be a better antenna than any existing tower type antenna and will be a satisfactory replacement for the Lualualei antenna. (U.S. Dept Navy, 1942b).


After receiving the positive results of the test the progress of construction picked up pace even further.

All through the valley terraces were leveled and made ready for the bomb-proof transmitter building, the ‘helix house’, living quarters, barracks, sewage plant, and system of roads. The ‘keying’ or actual message-sending from Haiku was to be done at Wahiawa, seventeen miles away. To make the connection, a cable was run to Kaneohe, and thence around the island by road (Woodbury 1946:362).


One of the components of the radio station was an elaborate copper wire ground system. The wires were to be attached to poles and set out in a grid pattern to cover the valley floor. The grid system worked to increase the radiating capacity of the system. In a letter to the Morrison Knudsen Company in 1943 Martin Broan stated that, “The ground system at this station is much more elaborate than most staions of this type have due to the high power and the poor conductivity of the soil in the valley (pg.5). The work was very difficult because the terrain was muddy, preventing the use of bulldozers, and covered with the heavy twisted hau trees. The holes for the poles were dug out by hand and dozens of wires, each a quarter of a mile long, were placed in a spider web fashion following the plan laid out by R.C.A.


The cable car systems were improved. On the North Ridge a sixty foot tower was constructed in order for the car to be able to pass over an overhanging ledge. The South Ridge had few snags. Once the car was stuck and two men were forced to spend the night in the dangling car. Another trolley system was placed along the ridge so that it would pass over the five anchor spots and materials could be brought in this manner. Workers remained on the mountain for days on a regular basis (Woodbury, 1946:365).

 Construction of the five main anchors was ready to begin.  [They] were lofty A-frames made of eight inch steel I-beams set down ten feet or more in solid blocks of concrete . . . Before they could be set, the gangs had to notch out the crest of the mountain with picks and shovels till solid bedrock was exposed, then blast a well to hold the anchor block. Concrete was mixed dry down at the Kaneohe batching plant and trucked in to the valley floor. Here it was watered and sent aloft in big steel buckets, than transferred to the [cable car] and poured directly from it through a portable chute (Woodbury 1946:365).

Finally the heavy steel wires were to be put in place. Several options for this were discussed including one from R.C.A. which was deemed unrealistic. The method used consisted of using a pulley in the valley to raise the loop of cable by gradually lifting the end of the loop off the ground (Figure 6). Weight boxes were built and placed behind the anchors on the ridge as counterweights to the fourteen ton pull of the cable. The weight boxes were each 14 tons. Woodbury relays the risk involved in this one process. One worker slipped while attempting to place one of the weight boxes and fell approximately 400 feet but was basically unscathed (Woodbury, 1946:366).

One last project was left that involved the arduous construction along the ridge: ” . . . a complete bomb-proof block house on the highest pinnacle of the ridge--at 2850 feet--to contain the ultra-highfrequency unit which could maintain communication with Kaneohe and Wahiawa if the regular cables failed” (Woodbury, 1946:367). This structure was referred to as the ‘CCL House’ or the Communications Control Link Station. While there were “direct lines for telephone and telegraph . . . connected to Pearl Harbor, Kaneohe Naval Radio Station, Ford Island Naval Air Station, Wahiawa Naval Radio Station, and Lualualei Naval Radio Station”, this equipment housed at the pinnacle of the ridge allowed for an “ultra high frequency direct beam radio that also could reach the points listed above” (Morrison, 1942: 4). With this, the radio station was complete and ready to handle high priority information during the war.

The CCL building, circa 2015. On routine maintenance days--before they were disallowed by the BWS--volunteers from the Friends of Haʻikū Stairs would remove trash and graffiti from buildings such as this
The CCL building, circa 2015. On routine maintenance days--before they were disallowed by the BWS--volunteers from the Friends of Haʻikū Stairs would remove trash and graffiti from buildings such as this

On the valley floor, not only had workers been building the bombproof transmitter building, but they also had completed an entire complex of other necessary structures to insure complete operational facilities for the radio station (Figure 7). From the original scope of work (Project 809, authourized by change A04, dated May 15, 1942, and Project 860, authorized by change M-4, dated June 19, 1942) projects included: 

” . . . preliminary survey work, and construction of a radio transmitting station, including equipment, services, etc., by R.C.A.; construction of a bombproof transmitter building, barracks for 35 men, roads, walks, fence, and services; clearing; power supply; setting poles for RF line; installation of water cooling system, and oil cooling system; telephone installation; construction of anchorages; and transportation and installation of eqipment. The estimated cost of the work was $1,198,000 . . . Additional work included construction of quarters for the Officer-in-Charge, addition of 600-KW. standby Diesel-electric generator, additional construction for TCG transmitter, testing aerial, addition of filter beds to septic system, and repair of fire damage. The estimated cost of additional work–after adjustment deductions for work omitted from contracat (because of delayed deliveries) and later performed by Construction Battalions– was $521,300" (U.S. Navy, 1944:A-816).

The installation used its own lumber yard within the valley for these projects (U.S. Dept. Navy, site map).

In November, 1942, there was a decision to locate the TCG (Transmitter, Coast Guard [U.S. Dept. Navy, 1944: A-811]) antenna that was scheduled to be placed at another Naval Radio Station, Lualualei, at the Haiku location instead. Preliminary specifications for the TCG antenna stated: