Real-time GPS Monitoring of Structures in the
Pacific Northwest

Howard Hansen Dam

Howard Hansen dam is a leaking earthen embankment structure located above Tacoma and nearby cities in the southern Puget Sound region of Washington State. The dam was completed in 1961 with the primary purpose of flood control and providing water to Tacoma and the Green River Valley. In January of 2009 the US Army Corp of Engineers (USACE) discovered seepage and two surface depressions in the right abutment of the dam, suggestive of internal erosion of the embankment by the seepage. Although the USACE does not believe the dam is at immediate risk of failing in any style similar to the Teton dam, concerns about its potential for failure have led to the injection of a grout curtain within the embankment, the drawdown of the resevoir levels behind the dam, and the installation of a host of new monitoring technologies, include high-precision GPS monitoring of 10 points around the dam operated by the Pacific Northwest Geodetic Array at Central Washington University.

Quantifiable position measurements of dam deformation are made possible by installing solid monuments with real-time GPS receivers across the dam. This study is intended to provide real-time position estimates across multiple transects of the dam and local geology to aid dam operators and engineers in hazard mitigation. The overarching goal is to determine typical deformation limits of the Howard Hansen dam caused by seasonal run-off/temperature changes, landslides and/or earthquakes. As a consequence, the applicability of real-time GPS data and experiment design will also be assessed.

Central Washington University (CWU) has partnered with the United States Army Corps. Of Engineers (USACE), Seattle Public Utilities (SPU) to instrument the Howard Hansen dam. Experiment design and GPS instrumentation was completed by SPU and CWU, the USACE provided the test site and constructed all the solid GPS monumentation located directly on the dam, CWU PANGA installed three solid bedrock reference sites and designed/constructed communications between all the receivers. This project was designed by employing the methods used in SPU's successful Tolt Dam monitoring prototype. Data processing and archiving will be completed at the PANGA laboratory at Central Washington University. The research staff of PANGA will provide professional expertise in GPS data interpretation and evaluation of monitoring techniques and methodology to the USACE, US Bureau of Reclamation and all others responsible for dam safety.

An array of 10 GNSS stations, known as nodes, are placed on various key features as identified by the USACE geology and dam safety representatives. The antennae, receivers, mounts, cables, and communications are provided on loan from CWU and WSRN partners for the duration of the study. The antennae are mounted on these key features in groups, with cables connected to four enclosures that house receivers, backup power, and communications equipment. Each enclosure sends data from these nodes via radio links to a main processing server/data logger. As part of this study it is further proposed that a Continuously Operating Reference Station (CORS) be established onsite that would become part of the National Geodetic Survey's National CORS program, providing a permanent tie to the National Spatial Reference System (NSRS) for use by USACE and as a public amenity for the region. All materials for the CORS are provided by CWU. The CORS will be located on the peak and be designated "HAHD". Several key features were identified at the site as desirable for structural study by the USACE geologist and dam safety staff. The CWU licensed geologist has also proposed several node sites valuable for geophysical studies relating dam structure and local geology of this site.

This "PEAK" node location will become the NGS National CORS site, and will be situated on an open bedrock section of the ridge about a mile to the northwest of the site. Central Washington University has secured permits with the DNR for use of this site and will install a short drill-brace type mount directly into the bedrock on the peak early this spring.

SPU's Tolt pilot study used the existing Washington State Reference Network as an active reference framework for hierarchical monitoring program involving a suite of motion engines to detect motion and relative positional changes over short and long baselines. This tiered approach enables detection of motion resultant from regional influences (i.e. tectonic plate movement) and more localized influences (e.g. pool level, landslides).

TWR-1 (above right). This tower node is mounted on a post securely bolted in three places to the northwest corner of the tower. The post extends 4ft above the top of the tower. 3.5in and 2in pipes and respective leveling mount types are employed.
BRDG-1 (above left). This bridge node is located on the tower access bridge over the vertical support nearest the tower. The mount is of the same type employed on the 520 floating bridge in Seattle, WA.

BDR-H. This short-drill brace node is one of two in the downstream bedrock area of the site. Located on the higher of the hummucks along the spillway, this site currently serves as the anchor node for the short-baseline motion engine studies. The mount is constructed in the same fashion as all hard rock GPS monuments employed as part of the Plate Boundary Observatory (PBO).


BDR-D node. This is the second of the two downstream bedrock nodes installed for solid bedrock reference (note site BDR-H on knoll in upper left distance). This area, adjacent to the spillway, incorporates both engineered dam materials and existing bedrock. The mount is also collocated with an existing survey cap.

The short-baselines afforded by this dense GPS network geometry allows sub-centimeter resolution at data rates of a sample of second. These GPS data, therefore, will provide instantaneous strain measurements of the Howard Hansen Dam during earthquakes, landslides, floods and regular seasonal temperature/rainwater run-off events -important to ascertain the nature of structural response and to discriminate between normal recoverable dam motion and irreparable damage.

FND-1 and FND-2. Above the grout wall, these two nodes will be driven post mounts just off the shoulder of the access road. 3.5in, Schdl-80 iron pipe is driven to a preferred depth of 5 to 6ft with 3 to 4ft exposed above the surface. A one foot piece of the same type pipe is welded on the top; this extension is threaded for a specialized leveling mount engineered for GPS antennas. Antenna cables from both (LMR-600, provided) run along the surface to GNSS receivers housed in a signal box and connected to the processing servers by radio. The location of the housing signal box is purposely located near existing electrical hand-holes on the same road shoulder for ease of AC power connections.

Results or reports from the CWU or SPU processing are solely based on interpretations of the raw data as probable positional changes and do not represent any conclusions as to actual dam safety conditions. CWU is simply studying a site and conditions for academic purposes only, and SPU is simply seeking to study the nature of structural movement as one of many studies that may possibly benefit future initiatives.

Ideal representative mount for all embankment installations (EMB-C, L & R) EMB-C. These are pile driven mounts at the center of the embankment just off the south shoulder of the access road. The preference was to drive the post in 1 to 2ft behind one of the existing bollards for protection and emplaced to ensure avoidance of future borehole locations (indicated on a plan provided by USACE). A LMR-600 antenna cable runs along the surface behind the bollards to a signal box as shown in map diagram below.
EMB-L & EMB-R: Added in the second phase, these nodes are mounted and connected in the same manner as EMB-C.

GPS antennae are mounted on key features in groups, with antenna cables connected to four enclosures that house the receivers, backup power, and communications equipment. Each enclosure sends data via radio links to a central processor. External communications links to CWU from the site will be handled by radio link from a nearby peak and will be independant of public communications grid. Installed auxillary power will also avoid data interuption during power outages expected to exist during an earthquake or other natural disaster.

Monitoring solutions for the Howard Hansen GNSS monitoring network with map view of node locations and data communication schema.