Sunday, January 31, 2010

FIRESIDE CHAT: Hit-and-Run Hudson River PCB Dredging

[A condensed version of this Fireside Chat was published in the Sunday Gazette newspaper (Schenectady, New York) on 31 January 2010.]

Phase I of the largest PCB dredging project ever undertaken began this past year on the Hudson River at Fort Edward. It went badly, with dredging completed in just over half of its planned area, and with discovery of a previously unknown, deeper layer of PCB-tainted sediments that threatens to extend the planned six-year project for years longer. Just how badly it went is a little-known fact, though hints may be gleaned from two new draft reports prepared by the U.S. Environmental Protection Agency (EPA) and General Electric Company (GE) to evaluate Phase I. GE reported that dredging mobilized nearly 25 times the expected amount of PCB-tainted sediments… but even that is probably just the tip of the iceberg.

The beginning of dredging evoked reactions ranging from celebration about the prospect of an eventually-clean Hudson to concern about health effects, water pollution, noise and diminished recreational river use. Phase I included the drama of PCB levels in water and air creeping upward, and eventually exceeding EPA limits, forcing GE to halt dredging on multiple occasions. The EPA Hudson River Office deserves credit for trying to balance conflicting imperatives forced upon it: to dredge, but to maintain acceptable dredging conditions. The office also was forthright in disclosing problems, and willing to share technical data.

My expertise in environmental toxicology and human health risk assessment, and experience consulting on power plant projects affected by dredging, motivated me to observe dredging operations, analyze data, and evaluate dredging impacts. My conclusions are drawn from findings such as these:

--PCB concentrations in water were grossly underestimated because they were measured miles downstream of dredge platforms, after massive dilution had occurred;

--Dredge buckets mobilized PCB-tainted sediments into the water by pounding the river bottom, and then closing their 15-foot wide jaws;

--Sediments were forced to billow upward, out of the open tops of dredge buckets and into the downstream river flow;

--Most mobilized sediment did not reach monitors miles downstream, and so did not get counted in estimates of PCB that was mobilized;

--Those mobilized sediments instead fell back to the river bottom, still mobile, but often outside computer-defined dredging 'prisms' and beyond the reach of future dredging;

--Sediments lifted above the river leaked back into the river from dredge buckets that were prevented from sealing by logs, stones, boards, metal, and concrete construction debris in the river bottom;

--Highly concentrated, highly toxic liquid PCBs were encountered that could not be retained by clamshell dredges; and

--As a result, the preponderance of dredge-disrupted sediment was mobilized rather than placed into waiting barges (I estimate 75 percent was mobilized by standard five-cubic-yard capacity dredge buckets; smaller buckets, though, also were used, in sensitive areas usually near shore).

These impacts went largely unrecorded and unquantified because dredging impacts in air and water were monitored inadequately in the 90 acres of river included within Phase 1. The Phase 1 area was divided into 18 work units, with just a single dredge platform in each, and no more than 12 platforms — usually fewer — active at any time. Of critical importance were measurement of waterborne and airborne PCB levels, and employees’ personal exposure to airborne PCBs. All of these were measured inadequately if at all.

As waterborne PCBs are the primary source of airborne PCBs over the river, waterborne PCBs should have been measured where dredging occurred — but this was not done. Instead of sampling water in the Phase 1 area near Rogers Island, water was sampled about five miles downstream. Thus waterborne PCB concentrations at dredging sites were anyone’s guess, determined by the also-unmeasured dilution factor that occurs as dredge-mobilized material travels five miles downstream. Similarly, airborne PCB concentrations at dredge sites were anyone’s guess, because airborne PCBs were not routinely measured at dredging platforms.

EPA assured me that GE had generated 243 air samples via monitors worn by employees. With 500 employees engaged for the project, however, over 30,000 samples would have been expected. Still, the 243 samples might have been useful to indicate airborne PCB concentrations, except that the monitors that were used had very high (occupational) detection thresholds and EPA was unable to indicate whether any of the 243 samples originated from the dredging corridor (as opposed to the treatment facility). My photographs of employees on a dredge platform revealed no personal air samplers.

Airborne PCBs were monitored by portable air samplers adjacent to dredge platforms, on opposite shores of the river. These samplers recorded 24-hour average airborne PCB concentrations, but three problems undermined their usefulness. First, leaking dredge buckets produced PCB-tainted sprays that employees could inhale, that dropped back to the river surface before reaching onshore monitors. Second, monitors collected air samples from all directions, massively diluting air coming from the particular direction of a dredge platform, so onshore monitors failed to capture actual airborne PCB levels over the water.

Each portable air monitor also was moved downstream with its neighboring dredge platform. No monitors remained to record evolving airborne PCB levels at any dredged location. Indeed, as the area of dredged river increased to its maximum, the concentration of airborne monitors per acre of dredged river declined, because no monitors were added to cover the increased area of dredge river.

Hence, the third problem: PCB entry into the air probably will reach steady rates long after portable air monitors were withdrawn downstream along with dredge platforms. Indeed, PCB release from the river surface probably would not reach steady rates until long after Phase I was completed. By then, portable monitors will have been withdrawn, not only downstream but altogether, in what I call ‘hit-and-run dredging’.

These and other observations lead me to conclude:

--Water monitoring remote from dredging is inadequate, so reported mobilization, nearly 25 times the anticipated amount, probably is still just the tip of the iceberg;

--Airborne PCB monitoring via portable onshore samplers is inadequate to quantify residents’ exposure;

--Personal monitoring of dredging personnel as implemented in Phase I is inadequate to quantify their exposure to airborne PCBs;

--Airborne PCB levels probably are higher than measured levels, and possibly unsafe; and

--Dredging has turned exposed people into experimental subjects, and may turn riverfront communities into the subjects of epidemiology studies that may continue for generations to come.

I recommend the following for a possible Phase II:

--Water and air monitoring, including personal monitoring, should occur together at dredge platforms;

--Permanent air and water samplers should be installed to confirm EPA safety claims and protect public and environmental health; and

--Dredging Project Phase I evaluation, aimed only at improving dredging in Phase II, also should consider termination of the project.


Copyright © 2010 by The Center for Health Risk Assessment and Management, a Division of RAM TRAC Corporation