Amendment 0002 2008 SSC Lock Investigation Report.pdf

STENNIS SPACE CENTER

HANCOCK COUNTY, MS.

NAVIGATION LOCK & BASCULE BRIDGE

LOCK INVESTIGATION REPORT

Prepared by: Prepared for:

Mobile District SEPTEMBER 2008

National Aeronautics and Space Administration, Stennis Space Center

Table of Contents

i. Executive Summary

1. Introduction and Statement of Problem

2. Scope.

3. References.

4. Project Location and Features

5. Inspections and History of Problems

6. Instrumentation

7. Hydrographic Survey

8. Subsurface Investigations

9. Subsurface Conditions

10. Analyses

11. Conclusions

12. Recommendations

Appendices A. Logs of Borings and Laboratory Test Data

B. Boring Location Map and Subsurface Profiles

C. Geotechnical Analyses

D. Structural Analyses

E. Alignment Pin Surveys

F. Hydrographic Survey

G. Statement of Work

H. ROM Cost Estimates i

Executive Summary

This report presents results of an investigation into cause of movements of navigation lock monoliths that have occurred at the Navigation Lock and Bascule Bridge project at the NASA Stennis Space Center (SSC) in Hancock County, Mississippi. Concern about the movements, associated damage to the lock structure, and possible adverse impact on future lock operations grew because of a large spall that developed in a lock monolith following the last lock unwatering in 1998-1999 and the possibility that the movement causing the spall was continuing. A concrete spall is a broken chunk of concrete. The spall includes parts of the top 3 concrete lifts and has approximate maximum visible dimensions of 5 ft width transverse to lock centerline, 13 ft length along lock, and 15 ft high.

The project includes a navigation lock, bascule bridge, canal, dike, and overflow spillway. The project was constructed 1963-1964. The lock is used for barge transportation of liquid propellant. The SSC mission to test large liquid propellant rocket engines is expected to expand. The lock is expected to play a more significant role in engine testing where large quantities of propellant will be transported to the test site by barge.

The investigation of the navigation lock was funded by NASA and conducted by the United States Army Corps of Engineers (USACE), Mobile District in 2007-2008. The investigation included review of existing information, 10 exploratory soil borings, laboratory testing, installation of 7 piezometers, review and analysis of instrumentation, and geotechnical and structural analyses. The purposes of the investigation was to determine the cause of observed displacements of the lock wall, to evaluate the suitability of the project for navigation in the future, and to provide recommendations for such remedial actions that may be appropriate to extend the functional life of the project for navigation.

The project is located in south Mississippi within a subsiding region that includes southern Louisiana. The lock is made of massive concrete monoliths. Most of the lock monoliths are founded on a layer of medium to dense sand. Below the sand layer at most locations is a layer of clay and silt that is highly compressible. The clay and silt layer contains decayed wood and organic soil in varying quantity up to about 35%, but averaging about 5% by weight. The thickness of the clay and silt layer varies from none at locations near the middle of the lock chamber to about 20 feet at locations near both ends of the lock chamber. The large concrete spall occurred in Monolith 20L, which is a chamber monolith on the left lock wall located near the downstream end of the chamber and just upstream of a chamber transition monolith (21L) and a lower valve monolith (22L). The thickness of the silt and clay layer is about 18 feet and is overlain by about 22 feet of medium to dense sand on which the 20L, 21L, and 22L monoliths are founded. A generally dense sand layer at least about 70 feet thick is located below the silt and clay stratum. This sand layer extends to the explored depth.

ii

The movement that likely caused the large concrete spall in monolith 20L following the lock unwatering in 1998-1999 was determined to be differential settling and tilting of monoliths 21L and 22L with the top of those monoliths moving away from the chamber approximately 0.15 feet. Most of the movement occurred over a period of about 2 years in Dec 1998 to Jan 2001 beginning during or soon following the 1998-1999 lock unwatering. Heave of approximately 0.05 ft normally occurs during unwatering of the lock and settlement of the same amount is normal after rewatering the lock. The tilting of 21L and 22L occurred during this period with the chamber side of the monoliths settling little or no more than the normal retracement of the heave associated with the lock unwatering. The land side of the monolith 21L and 22L settled more than the normal retracement of heave to cause the tilting. In addition to tilting away from the chamber, monolith 21L tilted slightly upstream but was restrained by contact with monolith 20L.

Compressive stress associated with this contact and frictional shear stresses caused by the horizontal movement of monolith 21L away from chamber likely caused the spall at monolith 20L.

A similar tilting movement away from the chamber was noticed in 1984 at monolith 23R on the right lock wall, following the first lock unwatering in 1982. That movement is suspected to have begun during or soon following the lock unwatering.

The cause or causes of the tilting movements at monoliths 21L and 22L were not reliably determined, but a probable cause was determined. The probable cause of the tilting movement is removal of soil, most likely into the foundation drainage system, and subsequent displacement of soil or the structure into the voids caused by the soil removal.

The rate and amount of displacement depends on the rate and amount of soil removal and can increase over time. The timing and other details of the tilting movements are generally consistent with this cause. The drainage system for the lock wall includes foundation drains located beneath the monoliths. The foundations drains are perforated corrugated metal pipes that are encapsulated within sand and gravel filter material. The foundation drains discharge to a collector pipe that is embedded within the monoliths and carries the discharge water to lower pool. Sand was observed in the discharge from the foundation drain system during lock unwaterings. It is unclear if the loss of foundation soil is occurring just during lock unwaterings or also while the upper pool is in the lock chamber. Discharge velocities of 6.95 and 7.75 ft/sec were measured by divers with upper pool in the chamber at the left and right foundation drain outfalls respectively, corresponding to flows of 1700 and 1900 gpm in the 10-inch diameter outfall pipes. The loss of foundation soil is likely to be occurring at a much smaller rate or not at all while lower pool is in the lock chamber because of much smaller expected drainage flow.

Possibly perforation sizes were too large in the foundation drain pipes and possibly foundation drain pipes have been broken by settlement. Either of these would allow entry of foundation soil into and soil removal by the foundation drain system. The tilting movements at 21L, 22L and 25R occurred or may have occurred during or soon following the lock unwaterings. Sinkholes have been observed at or in the vicinity of these monoliths during the lock unwatering in 1982 and at a borehole in 2007.

iii

There is substantial risk of further foundation soil removal and displacement of the structures. The expected amount of future displacement without remedial action can not be reliably predicted but could be larger than displacements experienced so far. Based on the history, large displacement from foundation soil removal is considered more likely to occur during or following future lock unwaterings than during normal lock operations.

The risk of damage to structures from this possible cause is expected to either increase or remain constant over time.

The risk of the most concern is if displacements become large enough to make the miter gates inoperable. Other events caused by excessive displacements of the monoliths could include more concrete spalling, broken pipes, broken waterstops, and associated leakage.

Because of more observed displacement in the past, displacements at the lower miter gate are of more concern than at the upper miter gate. The lower miter gate monoliths should be able to handle several inches of tilting movement away from the chamber. Tilting movements cause a warping of the gate and probable early formation of fatigue cracking.

The amount of movement that the miter gate can withstand will depend on the stiffness of the gate. Connecting elements of the gates can be adjusted as needed to accommodate movements.

To reduce the risk of structural damage and loss of future operability of the structure, remedial actions should focus on assuring that more foundation soil is not lost from under the monoliths. Grouting or otherwise plugging up the foundation drain system is the most reliable way to accomplish this; however, this action would adversely affect the foundation drainage and structural stability particularly while lock is unwatered. A pumped well dewatering system to be operated while the water in the lock is below lower pool during lock unwatering is recommended. Other recommendations are made to acquire more data to monitor the displacements and other conditions at the lock.

1. Introduction and Statement of Problem

The Stennis Space Center (SSC) Navigation Lock and Bascule Bridge project is owned by the National Aeronautical and Space Administration (NASA) and is operated by NASA contractors. The project was designed by the United States Army Corps of Engineers (USACE) Savannah District. The project was constructed 1963 to 1965 under the supervision of the USACE Mobile District. The lock is used for barge transportation of liquid propellant. As NASA’s mission continues to evolve, the SSC mission to test large liquid propellant rocket engines is expected to expand. The lock is expected to play a more significant role in engine testing where large quantities of propellant will be transported to the test site by barge.

The mass concrete gravity monoliths that comprise the lock have experienced displacements of varying magnitude since construction. Displacement is presently most noticeable at the downstream end of monolith 20L. Since 1998, a large concrete spall has developed in monolith 20L. The spall is irregularly shaped, including part of the top three lifts of concrete in 20L, and extends from the 20L-21L joint to a maximum observable distance about 12 feet upstream of the 20L-21L and from the land side vertical face to a maximum observable distance about 5 feet from the land side vertical face. The lock chamber side vertical face of monolith 20L has an apparent bulge at lifts two and three down from the top. Leakage of water from the lock was observed in the spall crack and the 20L-21L joint when the water level in the lock chamber approaches upper pool level and when upper pool level is sustained in the lock chamber. The condition at the monolith 20L location has worsened since 1998 to cause some concern for the lock’s future integrity. Relatively large displacements have also occurred at monoliths 26R and 6L, but these appear to be older displacements that have not worsened in recent years.

2. Scope.

In 2007, an investigation of the lock was funded by NASA and conducted by the USACE Mobile District. The statement of work for the investigation is included as Appendix G.

In summary, the statement of work included

• Thorough Periodic Inspection and Continuing Evaluation Survey (PICES) report;

• Thorough geotechnical/structural investigation to ascertain cause for the continual movement of the Lock monoliths on both lock walls;

• Determination of the particular cause for the movement of Monolith 20L; and

• Remediation plans for the Lock system, bridge and associated equipment as required for reliable operations.

The PICES inspection was made 13 Mar 2007. The PICES report is being provided as a separate document.

This report documents the geotechnical/structural investigation and includes conclusions regarding the causes of the movements of the lock walls and recommendations for actions that are appropriate to increase the reliability and extend the functional life of the project.

Deviations from the Appendix G scope included:

a. Ten borings were drilled. One boring planned to be drilled from the top of and through monolith 20L was omitted and three borings (SC-8-07, SC-9-07, and SC-10-07) were added. Piezometers were installed in 7 borings (SC-1-07 to SC-7-07).

3. References.

#1. “Construction Plans for Navigation Lock and Lock Water Supply, NASA Marshall Flight Space Center, Mississippi Test Facility, Mississippi”, USACE Mobile District, Oct 1963.

#2. “Inspection of National Space Laboratories, Lock & Bascule Bridge”, USACE Mobile District, Feb 1977, inspection date 14-15 Feb 1977.

#3. “Inspection and Evaluation of John C. Stennis Lock & Bascule Bridge, John C. Space Center near Bay St. Louis, MS”, USACE Mobile District, undated, inspection date 29-30 Oct 1991.

#4. “Navigation Lock & Bascule Bridge, Periodic Inspection, John C. Stennis Space Lock, John C. Stennis Space Center”, USACE Mobile District, Nov 1998, inspection date 17 Nov 1998.

#5. “Sea-level history of the Gulf of Mexico since the Last Glacial Maximum with implications for the melting history of the Laurentide Ice Sheet”, Alexander R. Simms, Kurt Lambeck, Anthony Purcell, John B. Anderson, and Antonio B. Rodriguez, Jan 2006.

#6. Contract Specifications for “Contract No. DA-01-076-ENG(NASA)1564, Navigation Lock and Water Supply at Mississippi Test Facility”, USACE, Oct 1963

#7. “NSTL Navigation Lock Lower Gate and Tainter Valves Maintenance”, Report No.

184-83-002, Stennis Space Center, 22 Jul 1983

4. Project Location and Features. The project is located at the Stennis Space Center in Hancock County, Mississippi on a tributary of the East Pearl River. The project includes a navigation lock, a bascule bridge located upstream of the lock, a canal, an overflow spillway, and an earthfill dike. The focus of this investigation is the navigation lock.

The lock has a 110 ft x 600 ft chamber and provides a nominal lift of 21 feet from (minimum) lower pool elevation -3 ft National Geodetic Vertical datum of 1929 (NGVD29) to upper pool elevation 18 ft NGVD29. The lower pool is influenced by river and tidal fluctuation and is most frequently at about el. +1 ft. The lock walls are comprised of mass concrete gravity monoliths. Monoliths 6L to 29L and 9R to 31R, including the entire lock except for parts of the guide and guard walls, bear directly on a sand foundation. Monoliths 1R to 3R of the upper guide wall and monoliths 32R to 37R of the lower guide wall are supported by piers that bear on timber piling. Monoliths 4R to 8R of the upper guide wall, monoliths 30R and 31R of the lower guide wall, and monoliths 1L to 5L comprising the upper guard wall bear directly on timber piling. (ref.

1, drawings 10/1, 10/4, 10/5 and 10/6)

A lock construction drawing (ref. contract drawing 10/1) indicates the vertical control to be based on MSL (Mean Sea Level). The Sea Level Datum of 1929 was commonly used in the mid 1960’s and was renamed to be the National Geodetic Vertical Datum of 1929 (NGVD29) in 1973. Accordingly NGVD29 is assumed to have been used for the lock construction. NGVD29 was largely superseded by NAVD88. Subsequent adjustments were made to NAVD88 including the datum of the last survey, which is NAVD88 (2004.65). The project is located within a subsiding region that includes southern Louisiana. The appropriate conversion at the project location from NGVD29 to the NAVD88 datum unfortunately is unknown. For this reason and unless otherwise indicated, elevations in this report are stated relative to the datum for which the respective elevations were originally recorded rather than to a consistent datum.

The wall monoliths on each side include a culvert that carries water from upper pool to the lock chamber while filling the lock and from the lock chamber to lower pool while emptying the lock. Water is discharged from the culvert into a system of six diffusers located across the channel at the lower gate sill. The culverts are comprised of 11 ft square holes at the chamber wall monoliths and vary in size at other locations. The concrete of the lower sill forms the floors of the diffusers. The roof and side walls of the diffusers are reinforced concrete slabs. Water is discharged from the diffusers to lower pool through side wall ports at elevation -29.00 to -26.37 ft NGVD29. (reference contract drawings 10/1 and 10/12)

A steel sheet pile wall is embedded into and extends below the lock chamber walls on both sides and upper and lower gate sills to form an enclosed partial seepage cutoff wall around the lock chamber. When the lock is filled with water to upper pool, the partial seepage cutoff wall impedes flow from the lock through the foundation to the pervious backfill adjacent to the lock. When the lock is unwatered, the partial seepage cutoff wall impedes flow from groundwater in the pervious backfill and the upper sand stratum through the foundation into the lock floor. When the water in the lock is at lower pool, there is only a small head difference between water in the lock and in the backfill. The bottom of the sheet pile wall is within a silt/clay stratum at most locations, but is within a sand stratum at borings S-4 and CS-4. The sheet pile wall penetrates to elevations of -56 and -61 ft NGVD29 at differing locations (reference contract drawing 10/3).

At and adjacent to the lock upper gate monoliths, steel sheet pile walls are extended in both directions perpendicular to the lock centerline from this enclosure and tie into embankments that include a core and core trench made of cohesive fill. The sheet piles and the cohesive fill core form a partial seepage cutoff that impedes flow of water from upper pool into the pervious backfill adjacent to the lock chamber walls. (ref. 1, drawing

10/3) Soil classified as CL (lean clay) and CH (fat clay) was considered cohesive soil suitable for the core and core trench fill.

Soil classified as SP (poorly graded sand), SM (silty sand) and SP-SM (silty sand) was considered pervious soil suitable for pervious backfill. Pervious backfill was used for the backfill located downstream of the core and core trench. (reference contract specification

pg. 2-2)

A foundation drain consisting of an 8-inch diameter, zinc-coated, perforated corrugated metal pipe (CMP) surrounded by filter material is located in the foundation adjacent to the sheet pile wall on both sides of the lock. At approximate 10-ft intervals along the wall, 8-inch CMP riser pipes connect the drain to a 10-inch CMP discharge pipe embedded in the lock wall monoliths at center-of-pipe elevation -22 ft NGVD29. The 10-inch discharge pipe carries drain outflow to lower pool downstream of the lower lock gate (reference contract drawing 10/3 and specification pg 21-1 to 21-3). Pipe elevation at point of discharge between diffusers at station 678A at monoliths 27L and 24R is at elevation -22 ft NGVD29.

Backfill wall drains consisting of filter material with a 12-inch diameter, bituminous-coated, perforated CMP are located on both sides of the lock in backfill adjacent to the lock monoliths. The CMP discharges to lower pool. The drains have 0.15% grades and have invert elevations of 1.0 ft NGVD29 at points of discharge into lower pool.

Backfill slope drains consisting of filter material with a 12-inch diameter perforated CMP are located on both sides of the lock in backfill adjacent to the lock excavation slope. The drains discharge to lower pool. The drains have 0.15% grades and have invert elevations of 1.0 ft NGVD29 at points of discharge into lower pool. The upstream ends of the drains have invert elevations of 2.45 and 2.33 ft NGVD29 respectively on the left and right sides of the lock. (ref. contract drawings drawings 7/5 and 7/6)

A lock floor drain system consisting of filter material with 8-inch perforated CMP pipe at 16 ft spacing is located just below the lock floor slab. These drains discharge into the lock chamber at elevation -20.0 ft NGVD29 from embedded pipes in the lock walls and from weep holes provided at the center of each floor slab on 16 ft on center each way.

The weep holes are gravel-filled 6-inch diameter holes through the floor slab. The top of the weep holes and the top of floor slab are at el. -22.0 ft NGVD29. (ref. 1, drawing 10/21). The weep hole gravel was specified to be ½ to ¾ inch in size.

Required perforation size in backfill wall drain and backfill slope drain CMP pipe was 3/8 inches. (ref. 1, drawings 7/5 and 7/6) The clear diameter of the perforations was specified to be not less than 75 percent of the perforations prior to bituminuous coating.

(ref. 6, pg 24-2) Requirement for perforation size in perforated pipe at the foundation drains and lock floor drains was not found.

The filter material to be used for the lock floor drains, backfill drains and all subdrains was specified to be a uniformly graded material graded between the following limits (reference contract specifications, pg 24-3):

Table 1: Gradation of Filter Material Specification in Percent Finer by Weight

Sieve Size From To

3/4 inch 100 100 1/2 inch 70 85

No. 4 55 70 No. 10 30 50 No. 40 0 10

The lock includes recesses for stoplogs to be used for lock unwatering. Stoplogs are not stored on site. Stoplogs were obtained from USACE Mobile District projects on the Black Warrior – Tombigbee river system for unwatering of the SSC lock in 1982 and 1998-99. The lock foundation was partially dewatered by gravity drainage using the foundation drain system during the two lock unwaterings. The lock foundation was dewatered using a concentric ring, multi-stage, pumped wellpoint system during construction of the lock.

A three-dimensional model was developed including the primary features of the lock and its foundation. Figure 1 is a view of this model looking upstream from the lower left bank showing most of the lock concrete, sheet piles, and the lower clay and silt foundation stratum into which the sheet piles are embedded.

Figure 1 - 3D Model of Lock Looking Upstream

Figure 1 Legend Olive – concrete Dark red – sheet piles Turquoise – clay and silt stratum

5. Inspections and History of Problems.

Construction phase information provided herein is from contract documents and Geotechnical Branch files archived on microfilm located at USACE Mobile District.

Inspection reports of the project by personnel from the USACE Mobile District prior to 2007 include trip report dated 1 Mar 1977 of a 14-15 Feb 1977 inspection, trip report dated 22 Jan 1985, and Periodic Inspection and Continuing Evaluation Survey (PICES) reports dated 29-30 Oct 1991, 17 Nov 1998, and 13 Mar 2007. A report (ref. 7) was made documenting the first lock unwatering for maintenance in 1982. Inspection reports earlier than Feb 1977 were not found. Most reports included mention of displacement and/or other indications of distress of portions of the lock walls at varying locations, which are summarized below.

Feb 1977. Approximately ½ inch of settlement at Monolith 6L was described and photographed in the report of the 14-15 Feb 1977 inspection (ref. 2, pg. 7, B-11, A-8, A- 17 and A-18). This estimate was of relative displacement based on observation of joint offsets. Sinks approximately 1.0 to 1.5 ft deep in backfill adjacent to lock monoliths 4L- 5L and 8L and bridge monolith #4 were photographed (ref. 2, pg. A-7 and A-8)

Nov – Dec 1982. The first of two lock unwaterings for maintenance was accomplished approximately 9 Nov 1982 to 11 Dec 1982 (Ref. 7). During inspection made 12 Nov

1982, there were bad leaks at joints at the lower gate monoliths (24L and 27R) on each side of the lock. A small amount of soil was being transported in the flow through these leaks. Midway through the unwatered period, two sinkholes of 2.5 cy each appeared on the outside of the south lock wall and one 1.5 cy hole was found outside the north lock wall. The sinkholes were backfilled with gravel and sand, and monitored. No significant change was observed as of 5 months later. The quantity of sand observed in the outflow from the left wall foundation drain was approximately twice the quantity from the right wall foundation drain. While the lock was unwatered, there was flooding of the Pearl River which caused the tailwater outside of the empty lock to be within an inch of the top of the stoplogs, or to approximate el. 7 ft. The water outside the lock remained high for a period of approximately one week.

Jan 1985. Approximately 0.5 to 1 inch of displacement were observed at monoliths 24R, 25R, 26R and 27R during the 22 Jan 1985 inspection. Movement of the lower right lock wall was noticed in the summer of 1984, but was believed to have possibly occurred sooner. The displacement was indicated to be a combination of settlement and rotation or tilting of the top of the wall toward the backfill. These estimates were of relative displacement based on visual observation of joint offsets.

Oct 1991. (Ref. 4, pg C-14 and C-17, and C-18) During inspection of 29-30 Oct 1991, seepage through the lowest visible lift line in monolith 20L was observed when the water level in the lock exceeded that level. The leakage was less than 1 gpm and had been noted previously in about the same condition. Settlement varying from ¼ to 1 inch was described at “3 of the 4” valve monoliths; or 10L, 22L and 25R. Based on joint offsets, Monolith 22L had displaced approximately 1 inch lower than monolith 23L at top of wall (Ref. 4, pg A-18, photo 34) and in the gallery (photo 36) and about 1 inch more horizontally away from the chamber (photo 37) in the gallery. Movements at joints 9L/10L, and 22L/23L and 25R/26R were recommended to be monitored.

The second lock unwatering for maintenance was accomplished approximately 9 Nov 1998 to 10 Feb 1999. A PICES inspection was conducted 17 Nov 1998 and alignment pins were surveyed 15 Dec 1998 while the lock was unwatered. Sand was observed by maintenance personnel in the discharge water from the foundation drain at the left wall during the lock unwatering. Sand was deposited near the foundation drain outfall at the diffuser area near the left wall. The sand deposit was removed while the lock was dewatered.

During the inspection of 17 Nov 1998, the top joint between monoliths 22R-23R had closed tight and had areas where spalled concrete had been repaired. Concrete had spalled at the top joint between monoliths 36R-37R. Water was draining into the lock chamber through the monolith joints at 23L-24L and 26R-27R, both at the level of the lock floor.

The flow was estimated to be about 25 gpm. Monolith 26R was approximately 1 inch lower than Monolith 27R and had shifted back horizontally about 1 ¼ inches, but this did not appear to be recent movement. Monolith 6L was ¾ inch lower than monolith 5L at their joint and the joint between monolith 6L and 7L had opened up to 1 ¼ inch. It was concluded that monolith 6L had rotated toward the upstream end of the lock.

A large concrete spall developed in monolith 20L in 1998 and its condition has since worsened. The spall is irregularly shaped, including part of the top three lifts of concrete in 20L, and extends from the 20L-21L joint to a maximum observable distance about 12 feet upstream of the 20L-21L and from the land side vertical face to a maximum observable distance about 5 feet from the land side vertical face. The lock chamber side vertical face of monolith 20L has an apparent bulge at lifts two and three down from the top. Leakage of water from the lock was observed in the spall crack and the 20L-21L joint when the water level in the lock chamber exceeds backfill grade (~el. 11 ft).

Relatively large displacements have also occurred at monoliths 26R and 6L, but these appear to be older displacements that have not worsened in recent years.

Another PICES inspection was conducted 13 Mar 2007. The largest displacements were observed at monoliths 7L, 20L to 23L, and 28R. Spalling of concrete and misalignment of guard rails was also observed, most notably and near at the monolith 20L/21L joint.

The most recent and worst spalling is at monolith 20L.

A sinkhole was observed 23 Jul 2007 at the location of boring SC-1U-07 in backfill adjacent to monolith 22L. The boring had been drilled 27-28 Jun 2007 and backfilled with sand soon thereafter. When this sinkhole was first observed, the borehole backfill had settled and left a neat 4-inch diameter hole looking like the borehole had not been backfilled. As of 7 Aug 2007, this sinkhole had irregular surface dimensions of approximately 30 inches by 30 inches. The hole increased in size below grade with its greatest dimension of estimated over 4 ft extending in a southwest trend. Depth at the center was approximately 18 inches and the maximum depth was over 24 inches. As of 5 Jun 2008, the sinkhole was described as being approximately 3 ft in diameter with maximum depth of 1.3 ft.

A diving inspection was conducted Jan-Feb 2008 by Bosarge Diving Company.

Velocities of 6.95 and 7.75 ft/sec and outflows of 1700 and 1900 gpm were measured at the left and right wall foundation drain 10-inch diameter outfalls respectively while upper pool was in the lock chamber. A deposit of sand was located near the left wall foundation drain outfall. It was undetermined if the sand came from the foundation drain outflow.

In summary, sporadic displacement and spalling of the lock walls have been observed throughout the project history. Excluding unreported but expected large displacements during and several years after construction due to primary consolidation of a clay stratum in the foundation, most of the larger displacements and spalling of the lock walls have occurred within a few years following lock unwatering periods. No large displacements were reported during lock unwatering periods. Sinkholes in backfill were observed until and including the first lock unwatering in Nov-Dec 1982, but have not been reported thereafter except for the sinkhole at boring SC-1U-07 first observed Jul 2007.

6. Instrumentation. Instrumentation at the Stennis Space Center navigation lock is comprised of a system of alignment pins on each lock wall, survey caps, and piezometers.

Offsets at monolith joints have occasionally been measured or visually estimated at locations with relatively large offsets. Locations of the instrumentation are shown on Plate F-1 – Boring and Instrumentation Location Plan of Appendix B.

Horizontal distances are measured between survey caps to monitor horizontal movements. The caps include four brass caps and four plastic caps. The brass caps are located on top of monoliths 1R, 27R, 8L and 24L and were installed in 1966. The brass caps were identified in the last survey by their lock station and offset from lock centerline locations (“0+52.00B, 83.00’ SOUTH” at monolith 8L, “0+52.00B, 83.00’ NORTH” at monolith 11R, “6+70.00A, 83.00’ SOUTH” at monolith 24L, and “6+70.00B, 83.00’ NORTH” at monolith 27R). The four plastic caps were identified in the last survey as “A”, “B”, “C”, and “D“. Caps “A” and “B” are located on ground approximately colinear with and are measured in conjunction with alignment pins 25R1 and 22L1. Caps “C” and “D” are located on ground approximately colinear with and are measured in conjunction with alignment pins 12R2 and 9L2.

Ten open standpipe type piezometers have been installed to monitor pore pressures in the backfill adjacent to the lock and uplift pressure in the lock foundation. Three piezometers, identified as PZ-1-00, PZ-2-00, and PZ-3-00 were installed adjacent to the left lock wall Sep 2000 by the USACE Mobile District. The other 7 piezometers, identified as SC-1-07 to SC-7-07, were installed May – Jun 2007 by the USACE Mobile District.

Instrumentation charts and evaluation of the instrumentation data are included in Appendix E.

7. Hydrographic Survey. A hydrographic survey of the lock chamber and channel areas near the lock was made 24 Jul 2007 09:33 to 13:45 by John Chance Land Surveys. The survey was made using a single beam system to survey water bottoms. The system consisted of a boat equipped with StarFix for positioning and HydroTRAC for obtaining water depths. Bar check was performed at 8 feet. Bottom of water elevations were computed from the water depths and water surface elevations. Water surface elevations for the upper and lower pools were obtained with real time kinematic (RTK) method with control at monument SSC CRSP.

Two digital spot elevation data files (“water depths.csv” and “rtk_points.csv”) containing coordinates of hydrographic survey points and some points on land respectively were included in files provided by John Chance Land Surveys to the USACE Mobile District.

The x,y coordinates in these files were in survey feet relative to the NAD83 state plane coordinate system, Mississippi East zone. The elevations in these files were in feet relative to NAVD88(2004.65) datum, using elevation 28.22 ft at monument SSC CRSP.

Upper pool was measured at elevation 17.678 ft NAVD88(2004.65) as of 7/24/2007 9:37.

Lower pool was measured at elevation 1.687 ft NAVD88(2004.65) as of 7/24/2007 13:12. The spot elevation data from the hydrographic survey point file (“water depths.csv”) was overlaid on maps showing the project structures and channels for presentation, as shown in Appendix F.

The design elevation of the top of the lock floor slab is -22.0 ft NGVD29. Most of the lock wall monoliths have design grades that slope from elevation -20.0 to -22.0 ft NGVD29 and protrude 7.5 ft into the chamber. The elevations of water bottom within the lock floor slab area, as indicated in subset of points excluding the 7.5 ft wide sloped areas, varied from -24.3 to -21.8, averaging -23.4 ft NAVD88(2004.65).

The survey was intended to be used for evaluation of possible scoured area and for comparison with future hydrographic surveys of the same areas to detect areas of erosion.

The survey was not made with enough resolution to be of much use for those purposes, particularly in the vicinity of the channel downstream of the lower sill where the design finished grade changes over just a few feet (reference sheet S-5 of Appendix F).Differences up to approximately 12 ft between the surveyed and the design finished grades are indicated in this area. Subsequent inspection of the lower sill area by divers in Feb 2008 indicated that not more than an inch or two of sediment existed except at a small area near the left foundation drain outfall. No holes were present in the diffuser roof slabs, which might be erroneously interpreted from the survey.

8. Subsurface Investigations.

8.1 Design Phase Subsurface Investigation. A subsurface investigation was conducted for the design of the project by USACE Mobile District in 1962 to1963. Boring logs and summary of laboratory testing from that investigation have been found in documents from USACE Mobile District and SSC files. Relevant information from these borings is shown on the subsurface profiles in Appendix B. The investigation included splitspooned borings on approximate 340 ft spacing along the left lock wall and on approximate 170 ft spacing along the right lock wall. Sampling spacing varied from continuous to 5 ft intervals between samples. Standard Penetration test (SPT) blow counts were recorded and samples were visually classified in accordance with the Unified Soil Classification system. A subsurface investigation report or other narrative describing subsurface conditions was not found. A summary of laboratory testing performed Jan 1963 by USACE South Atlantic Division laboratory for this investigation is included in Appendix C. This summary includes soil classification, void ratio, and specific gravity, horizontal permeability, and vertical permeability data.

8.2 Construction Phase Subsurface Investigation and Changed Condition. Finer-grained, softer soil than indicated in the design phase borings was encountered at some locations of the lock foundation subgrade during excavations for lock construction.

Construction work was suspended and 80 borings were made to delineate areas of unsuitable materials. Locations and drilling logs for these borings have not been found.

The softer soil, where present at shallow depth beneath foundation grade, was removed and replaced with compacted select pervious fill. The deeper excavations required unwatering to a lower level to meet the unchanged requirement that the groundwater level be lowered to at least 1 ft below the deepest part of the excavation. Representatives of the contractor Morrison-Knudson Inc. in a letter dated 1 May 1964 described the soil indicated in recently acquired borings at the lock floor area as indicating “a considerable predominance of clayey sand and sandy clayey soil”. The USACE Mobile District in a document dated 8 Oct 1965 described the softer soil as “soft clay occurring in lenses and vertical stringers” and that it appeared “that an old river channel crosses the lock chamber, so situated such that none of the original borings intercepted it”. A contract modification to revise the lock foundation was issued in amount of $925,006 and time extension of 77 days due to changed conditions. A draft document dated Apr 1964 from USACE Mobile District identified excavation grades to be changed as spot excavation from el. -

34.0 to -40.0 ft at monolith 11R; from el. -42.0 to -45.0 ft over full extent at monoliths 12R to 22R except that deeper excavation to perhaps el. -49.1 ft at monolith 13R and to perhaps el. -49.8 ft at monolith 18R might be needed pending additional investigation;

no change at monoliths 23R to 25R except for backslopes at 23R required for deeper excavation at 22R; spot excavation from el. -46.1 to -47.4 ft at monoliths 9L and 10L;

spot excavations from -36.8 to -38.7 ft at monoliths 12L and 13L; no changes at monoliths 14L to 17L; and spot excavations to -35.5 ft at monoliths 19L and 20L. The document also indicated that hard silt (ML) was present at el. -45.5 to -49.0 ft at monolith 18L and that additional investigation was needed to determine if it was unsuitable. The excavation grades that were finally established and excavated may have differed from those indicated in the Apr 1964 draft document and are unknown.

8.3 Recent Subsurface Investigation. A subsurface investigation was conducted for this study May – Jun 2007 by USACE Mobile District to supplement the subsurface information from the design phase subsurface investigation. This investigation included 10 splitspooned borings, laboratory testing, and installation of 7 piezometers. Logs of the borings and laboratory test reports are included in Appendix A. Boring locations are shown in Appendix B. Borings were continuously splitspooned to depths varying from 60 to 150 ft. Standard Penetration test (SPT) blow counts were recorded. A manual hammer was used for borings SC-1-07 and SC-4-07. An automatic hammer was used for borings SC-2-07, SC-3-07, SC-5-07, SC-6-07, SC-7-07, SC-8-07, SC-9-07, and SC-10-07.

Samples were visually classified in accordance with the Unified Soil Classification system. Laboratory testing included grain size distribution, Atterberg limits, moisture content, consolidation, and triaxial consolidated undrained compression with pore-pressure measurements (R-bar).

9. Subsurface Conditions.

Outside of the excavation and backfill for the lock, the surficial soil stratum is an upper silt and clay stratum. Underlying the upper silt and clay stratum and comprising the foundation material for the lock is an upper sand stratum. The upper sand stratum at most locations is underlain by a lower clay and silt stratum. A lower sand stratum underlies the clay and silt stratum where the lower clay and silt stratum is present. At some locations including at borings S-4 and CS-4, the lower silt and clay stratum is not present and the upper and lower sand strata join to form a connected sand unit.

The upper sand stratum is comprised predominately of poorly graded sand (SP) with some silty sand (SM). Occasional lenses of silt and clay are found at some locations. The design phase borings indicated the upper sand to be almost entirely very dense poorly sand (SP) with 100+ SPT N values; however, a changed condition was determined during construction. Borings and observations made during construction indicated that previously undetected lenses and vertical stringers of soft clay were present at some locations. The soft clay and clayey soil was removed and replaced with compacted pervious fill where detected at shallow depth under lock monoliths. Recent borings indicate that the upper sand is significantly looser than the design phase borings indicated and is predominately medium to very dense in relative density.

The clay and silt stratum is comprised of highly plastic clay (CH), lean clay (CL), highly plastic silt (MH) and contains wood and other organic matter in varying quantity at most locations. The organic content as measured by loss on ignition (LOI) laboratory tests varied from 0.2% to 35.1%, averaging 5.6% in the tested samples from the recent subsurface investigation. This average is biased to the high side because nearly all of the samples that were selected for the LOI tests were described as containing organic material. Where present, the bottom of the lower silt and clay stratum is located at approximate el. -69 +/- 2 ft. The clay and silt stratum varies from zero to approximately 21 ft in thickness, averaging about 14 ft thick.

For characterizing the subsurface conditions at locations where the lower silt and clay stratum is not present, elevation -70 ft was considered to be the bound between the upper and lower sand strata.

The lower sand stratum is comprised predominately of poorly graded sand (SP) with some silty sand (SM) and gravel. Occasional lenses of silt and clay are found at some locations. The lower sand was predominately very dense with 100+ SPT N values in the original design borings. Recent borings indicate that the lower sand is predominately very dense, but is slightly looser with SPT N values typically in range of 50 to 100+.

Pervious backfill materials were obtained from satisfactory material from required excavations for the project and from borrow areas. Recent borings in the backfill indicate that it is predominately poorly graded sand (SP) and silty sand (SP-SM) and varies in relative density from very loose to very dense, and is predominately loose. SPT N-values vary from zero to 51 blows per foot, averaging about 8 blows per ft. The very loose backfill material was found mostly in corners formed by lock monoliths of differing width.

Cohesive backfill materials were obtained from satisfactory material from required excavations for the project and from borrow areas. One boring (SC-8-07) penetrated the cohesive fill zone near its downstream limit. It indicated very soft (zero blow-count) clay (CL) and loose (N=6 to 7 blows/ft) silty sand (SM) at that location.

Wood was identified in samples at elevations varying from approximately el. -50 to -69 ft, or -15 to -21 meters, at the project site. Wood and other organic material at considerable depth has been found at locations around the northwestern rim of the Gulf of Mexico from Corpus Cristi, Texas to Mobile Bay, Alabama. Samples of organic materials at varying depth have been carbon dated (Table 1 of ref. 5) and a correlation can be developed between the depth and age of the materials. The data suggests that the lowest sea level in the Gulf of Mexico during the Last Glacial Maximum period was approximately 58 meters below current sea level about 13,000 years ago and has been generally rising to current sea level since then. Furthermore the data suggests a sea level at 15 meters below current sea level approximately 6000 to 8000 years ago and at 21 meters below current sea level about 8000 to 9000 years ago. Accordingly it is likely that the materials at elevations -50 to -69 ft, including the lower clay and silt stratum, are at least approximately 6000 to 9000 years old, or possibly considerably older if deposited during a previous ice age with low sea level.

10. Analyses

10.1 Structural Analyses.

A stability analysis was performed for a typical chamber monolith and a lower gate monolith. The results indicate that the monoliths are stabile against rotation and sliding during all load cases examined. A summary of results follow:

Assumed Design Values:

Saturated soil density = 125 pcf Moist density = 100 pcf Concrete density = 150 pcf Pervious fill friction angle = 30 degrees (no cohesion) Coefficient of sliding friction = 0.55 Maximum ground water elevation = measured plus 2 feet

CHAMBER WALL

Load Case 1: Normal Low Pool in Lock

Maximum bearing stress = 2.64 t/ft2 Minimum bearing stress = 0.5 t/ft2 Sliding safety factor = 1.68 (neglecting concrete floor)

Load Case 2: Lock de-watered With Saturated Fill

Maximum bearing = 1.95 t/ft2 Minimum bearing stress = 1.34 t/ft2 Sliding Safety factor = 1.8 (neglecting concrete floor)

Flood Condition: Five stop logs are available and are typically the most used.

5x4 = 20 ft height Sill elevation = -18 20-18 = +2 (this elevation is the same as the measured ground water elevation used in load case 2)

Load Case 3: Upper Pool in Chamber

Maximum bearing = 3.30 t/ft2 Minimum bearing = 1.55 t/ft2 Sliding factor of safety = 5.41

LOWER MITER GATE MONOLITH

Load Case 1: Normal Low Pool in Lock

Maximum bearing stress = 3.23 t/ft2 Minimum bearing stress = 0.9 t/ft2 Sliding safety factor = 2.7 (neglecting concrete floor)

Load Case 2: Lock de-watered With Saturated Fill

Maximum bearing = 4.48 t/ft2 Minimum bearing stress = 1.69 t/ft2 Sliding Safety factor = 3.43 (neglecting concrete floor)

Load Case 3: Normal Upper Pool in Lock

Maximum bearing = 3.23 t/ft2 Minimum bearing stress = 0.96 t/ft2

10.2 Geotechnical Analyses

Two-dimensional finite element seepage and displacement calculations at lock chamber monoliths 20L and 16L were performed using Geostudio 2007 software. These locations were chosen so as to analyze locations both where the lower clay and silt stratum is present and where it is not. Potential for liquefaction of sand foundation and backfill was analyzed by a spreadsheet method. Methods, parameters, assumptions, results, and conclusions from these analyses are described in detail in Appendix C.

11. Conclusions

#1. Monolith Displacements. The largest horizontal movements of the lower left wall since 1994 when monitoring of alignment pins began included differential settling and tilting of monoliths 21L and 22L with the tops of these monoliths moving approximately

0.15 ft (+/- about 0.02 ft) away from the chamber. Most but not all of the slipping between monoliths occurred at the 20L/21L and 22L/23L joints. Most of the horizontal movement occurred over a period of about 2 years in Dec 1998 to Jan 2001, approximately as shown in Figure 9 of Appendix E, beginning during or soon following the 1998-1999 lock unwatering. Relatively large heave of approximately 0.05 ft normally occurs during unwatering of the lock and settlement of the same amount is normal after refilling the lock. The tilting of 21L and 22L occurred during this period with the chamber side of the monoliths settling little or no more than the normal retracement of the heave associated with the lock unwatering. Settling of more than the normal retracement of heave did occur at the land side of the monolith 21L and 22L following the lock unwatering to cause the tilting. In addition to tilting away from the chamber, monolith 21L tilted slightly upstream but was restrained by contact with 20L.

Compressive stress associated with this contact and frictional shear stresses caused by the horizontal movement of 21L away from chamber likely caused the large spall at the land side and top of monolith 20L. Horizontal movement of the top of monolith 20L was relatively small and erratic. The largest and most sustained horizontal movement of 20L was approximately 0.04 ft toward the chamber and occurred from 2001 to 2004, approximately as shown in Figure 9 of Appendix E. Relatively small settling and horizontal movement at monolith 23L likely also occurred.

The first reported movements of the lower left lock wall monoliths was Monolith 22L.

Monolith 22L likely settled approximately 0.08 ft and tilted with top moving away from the chamber prior to Oct 1991 approximately 0.08 ft, based on the 22L/23L joint offset photographed in the Oct 1991 inspection. Although the time of the 22L movement indicated by Oct 1991 photographs is unknown, movement during or soon following the first lock unwatering in 1982 is suspected. Monolith 22L likely tilted about 0.02 to 0.04 ft more away from the chamber about from Jul 2001 to June 2007, based on analyses of alignment pin horizontal measurements and increasing distance measurements between pins 22L1 and 25R1.

Monolith 25R on which end-of-line pin 25R1 is located likely settled approximately 0.04 ft and tilted with top moving away from the chamber prior to Oct 1991 approximately

0.06 ft, based on the 25R/26R joint offset reported and photographed in the Oct 1991 inspection. The movement of 25R was first noticed the summer of 1984, but may have occurred earlier. Movement during or soon following the first lock unwatering in 1982 is suspected.

#2. Cause of Displacements. Some displacement of monoliths in reaction to stress changes in foundation is normal and to be expected. Displacements are generally considered normal if they do not significantly exceed design or otherwise tolerable amounts and are caused by expected fluctuations in site conditions, typically including fluctuations of the water level in the lock chamber, groundwater levels in the…