By Martin Teal, Mark Baker, Robert Ettema, and John Trojanowski
In the early morning of March 14, 2019, the Spencer Dam on the Niobrara River in northern Nebraska suddenly failed during a major ice run on the river. The embankment portion of the dam failed in two locations: the north breach about 650 feet wide, and the south breach was about 800 feet wide. Flow of ice and water through the breaches inundated areas downstream, washing away structures and leading to a fatality.
Following the failure of Spencer Dam, the Chief Engineer of the Nebraska Dam Safety Program (NebDSP) contacted the dam owner, Nebraska Public Power District (NPPD), who agreed that the failure should be investigated by the Association of State Dam Safety Officials (ASDSO), a national non-profit organization serving state dam safety programs and the broader dam safety community. ASDSO convened an oversight group and selected four members for an investigation panel. To ensure independence, neither the NebDSP nor the NPPD had any input on the selection of the Panel’s leader or members. The failure
investigation focused on identifying what happened (the physical causes of the dam’s failure), why it happened (the human and organizational causes), and the lessons learned from the failure (to keep such failures from happening again). The final investigation report was publicly released on April 21, 2020.
Spencer Dam was located on the Niobrara River in northern central Nebraska (NE) about 5 miles southeast of the town of Spencer, NE and about 224 miles northwest of Omaha, NE. The dam was constructed in 1927 to produce hydropower by means of a run-of-river project that had a relatively small reservoir which had become partially filled with sediment. Spencer Dam consisted of a 3,200 foot-long earth embankment (herein termed dike) that extended from one side of the river’s floodplain to the other side, was founded primarily on river sediment, and arced northeastward, directing flow along the embankment and toward the dam’s powerhouse and spillway (Figure 1). The multi-bay buttress-type concrete spillway structure had one ice/trash bay with a 10-foot wide lift gate, four 33.5-foot wide Tainter gates, and five 33.5-foot-wide needle-beam stoplog bays. Each needle beam stoplog bay contained five needle beams and six stacks of wood stoplogs. The needle beams were operated by the positioning of a pneumatic jack between a welded bracket on the side of the needle beam and the operator walkway. A hydroelectric powerhouse was also located adjacent to and north of the spillway, containing two turbines with a total capacity of 3 megawatts.
In general, the Panel conducted the investigation following the methodology stipulated in the ASDSO Dam Failure Investigation Guideline and drew from the experience of the Independent Forensic Team for the Oroville Dam Spillway Incident of 2017. Additionally, the Panel reviewed a large amount of data provided by both NebDSP and NPPD, conducted a site visit, and gathered and reviewed data from local, state, and federal agencies. The Panel also conducted in-person and telephone interviews with many of these agencies and with individuals who had insights to contribute.
The Panel’s effort to reconstruct the events of March 13 and 14 were hampered by a lack of first-hand accounts due to the remoteness of the site, the evening and early morning timing of the failure, and severe weather conditions during the failure. The dam’s operators were able to provide descriptions of what they saw during the event, but only at specific locations and times, and they were limited by visibility. Given the lack of first-hand accounts, the report describes the range of what might have happened and details the Panel’s opinion of the most likely scenario for the dam’s failure as described in the following section.
Based on the operators’ accounts, the evidence left after the failure, and other observations and data, the Panel found that the most likely failure scenario is as follows:
- A wet autumn and colder than normal winter produced frozen ground, substantial thicknesses of river ice cover and snowpack. A winter storm, characterized as a bomb cyclone, affecting the entire Great Plains began around March 12 and initially produced temperatures above freezing resulting in rainfall or mixed rain and sleet on the snowpack and frozen ground. This storm produced flooding and dynamic breakup of the river’s ice cover. As time progressed, weather conditions became colder and windier.
- During the evening of March 13, dam operators opened all four of the dam’s Tainter gates to their maximum six-foot opening on the spillway crest. They later released stoplogs from some of the other bays to increase outflow but were not able to open most of the stoplog bays due to ice-encrustation.
- Around midnight on March 13, a major ice run came down the Niobrara River, failing the Stuart-Naper Bridge and damaging the Highway 11 (Butte) Bridge, both located upstream from Spencer Dam (Figure 2).
- One or more ice jams occurred upstream from the dam, backed up flood waters and then burst sending a great amount of ice rubble and flood water toward the dam.
- 5. Ice rubble and flood water rapidly filled the reservoir, causing the reservoir to rise to the dike crest.
- Continued inflow of ice and water into the reservoir pushed some ice rubble over the crest and downstream slope of the dike. Ice pushed through the upstream brick wall of the powerhouse (Figure 3) and may have clogged the spillway gates.
- Flow overtopped the dike, causing the downstream side of the dike to erode. The erosion led to headcuts, which grew in several locations along the dike’s downstream slope. The dam’s embankment dike breached in two locations, the first breach occurring around 5:15AM. The breaches widened and discharged water and ice rubble downstream.
- The flow of water and ice failed the dam and swept through a house and other buildings located immediately downstream from the dam, causing their destruction and the disappearance of the lone resident (who was later declared dead by drowning). The flow spread over the channel and its floodplain downstream of the dam and was impeded by the approach embankment of Highway 281, located a short distance downstream of the dam. When the flow breached the highway embankment, it formed a major new channel through the breach.
- The ice run carrying ice and debris continued downstream, where several other bridges were damaged or destroyed. The Panel completed hydraulic modeling of the river downstream. The Panel concluded that the failure of the dam did not exacerbate flooding more than a few miles downstream and certainly not in the village of Niobrara 39 miles downstream. The following factors that led to this conclusion: the small size of the Spencer Dam reservoir, the several bridges and other restrictions that potentially caused ice jams, the massive size of the flood and ice run, and the decrease in peak flow (attenuation) of flood water as it traveled downstream in the wide river floodplain.
The flood of water and ice greatly exceeded the capacity of the dam and its spillways. In the panel’s opinion, there was nothing the operators at the dam could have done the morning of the flood that would have kept the dam from failing given the magnitude of the flood and ice run.
If the dam had not been present, the Panel believes that the structures immediately downstream would have not been safe during this flood of water and ice; and the highway bridge and the local structures including the house would likely have been washed downstream by the initial surge of water and ice. If the dam had been modified prior to the event to pass the flood and ice run, the downstream highway embankment would likely still have backed up water and ice, flooding and damaging the house, before failing the highway embankment.
The Panel identified two key, human factors contributing to the dam failure and consequences:
1. There is a notable lack of knowledge about ice-run-related potential failure modes generally in the dam safety industry. Specifically, NebDSP did not know that Spencer Dam had previously failed and was damaged in ice run events. NPPD had limited knowledge of past ice run events at the dam.
ASDSO maintains a database of 380 dam failures. Although the database is weighted toward more recent failures (post 2010), no dam in that database was reported to have failed during an ice run. The National Performance of Dams Program lists one dam failure due to an ice flow in 1976. The Dam Safety Industry generally lacks knowledge of how ice runs can impact the safety of dams in cold weather regions. Current dam safety best practices do not include evaluating run-of-the river dams for stability during ice runs.
Ice was involved with the 1935 failure of Spencer Dam. In 1960 and again in 1966, the dam’s gates and powerhouse were damaged by ice. These incidents do not appear in ASDSO’s database as ice-related failures. There was no consolidated history of the dam, and important records were lost, unorganized, or unavailable. While the dam appeared to be well maintained, no provisions were made to pass or prepare for ice run events. Furthermore, NebDSP predominantly relied on its dam inspection program to bring dam safety issues to the attention of the dam owner; latent vulnerabilities such as performance during ice runs floods are not addressed in the state’s inspection reports.
2. NebDSP and NPPD underestimated the potential of the dam to cause life-threatening flooding at the downstream house and property in the event of dam failure.
There was a lack of recognition that the house, Strawbale Saloon and RV campground situated just downstream from the dam would be at risk if the dam failed. One reason is that the Downstream Hazard Potential Classification (DHC) for the dam was “significant” when, in the panel’s opinion, it should have been “high.” Its Significant DHC rating resulted in less dam safety regulation including no requirement for an Emergency Action Plan (EAP). If the dam were designated a “High” hazard potential dam, there would have been a requirement for an EAP and there might have been a requirement to modify the dam to increase flood handling capacity.
The failure of Spencer Dam was a tragedy. One of the most important results of the investigation is the lessons that engineers and dam professionals can learn from the failure:
- Engineers working on dams, bridges and other infrastructure facilities at rivers in cold-weather regions need to assess whether the rivers are susceptible to periodic severe ice runs. If this susceptibility exists, it should be addressed in design. Dam facilities should be designed to be operated safely during these extreme weather events. Warning systems are one potential measure to reduce risk where ice runs form.
- More research needs to be done on the dynamic nature of rivers in cold weather regions, including weather systems in such regions, ice run formation, frequency, movement, damage, and how infrastructure like dams should be designed, maintained and operated to withstand ice run loading.
- Dam inspections, while valuable, are not adequate dam safety evaluations in themselves. Evaluations must include review of critical documentation and records. Potential Failure Modes Analysis at an appropriate level should be conducted as part of a dam safety review. Once the potential failure modes (PFM) are understood, inspection checklists should be modified to identify signs these PFMs are developing and/or the dam is vulnerable to them.
- Dam owners should maintain a complete and organized set of electronic records for their dam(s). A concise history of the dam with reference to key records and past incidents is invaluable.
- One of the most important responsibilities dam safety regulators have is to periodically assess the areas downstream of low and significant hazard dams to evaluate whether the hazard classification is appropriate. Documented formal procedures (including reviewing data such as aerial or satellite photography and verifying during the site inspection) should be adopted.
- For dams with people at risk downstream, Emergency Action Plans should be developed and exercised.
- Dams should have operation plans that include operations during extreme events.
The Panel’s hope is that these lessons resulting from the investigation of the failure, including natural events and human factors, will be taken to heart by the community of practice resulting in safer and more sustainable infrastructure for the public.
The Panel gratefully acknowledges the contributions of ASDSO and its Spencer Dam oversight group (Roger Adams, PE, Lori Spragens, Mark Ogden, PE, Dusty Myers, PE, Greg Paxson, PE and John France, PE), Irfan Alvi, P.E. (human factors), James Pawloski, P.E. (state dam safety regulatory issues).
Full text of the report may be found at https://damsafety.org/SpencerDamReport.
Martin Teal, PE, PH, D.WRE. is Senior Vice President with WEST Consultants (hydrology and hydraulics).
Mark E. Baker, PE was the panel leader and is Principal, DamCrest Consulting (dam safety programs and human factors)
Robert Ettema, PE, PhD is a Professor in the Department of Civil & Environmental Engineering, College of Engineering, Colorado State University (ice and hydraulic structures)
John Trojanowski, PE is President, Trojanowski Dam Engineering (hydraulic structures, concrete dams)