Comments on the Proposed Reissuance of the U.S. Army Corps of Engineers' Nationwide Permits
Comments on the Proposed Reissuance of the US Army Corps of Engineers' Nationwide Permits
About EWG's Nationwide Permitting Data
Comments on the Proposed Reissuance of the U.S. Army Corps of Engineers' Nationwide Permits: About EWG's Nationwide Permitting Data
The Corps' RAMS database is designed primarily as a workload tracking and management system for Corps staff. Corps regulatory personnel use RAMS to estimate the number of staff hours devoted to processing permits, for simple correspondence with parties to a permit action, for notes about a permit's location and size, for tracking of the dates and times of steps in the permit process, and for a variety of other purposes. Many district offices use RAMS for quarterly regulatory reports that are forwarded to Corps headquarters in Washington, DC. Headquarters staff use these quarterly reports to monitor staffing and workload requirements of the district offices, and to evaluate the performance of Corps district offices. Districts that use RAMS for quarterly reporting generally enter information on all activities authorized under Corps permits of which they are aware.
Although RAMS does represent the most comprehensive database on regulatory actions taken under Section 404 by the Corps of Engineers, it is not, and should not be considered, a comprehensive list of all actions that the Corps has authorized. Some Corps districts have taken pains to back-enter Section 404 permits that were issued before the district began using RAMS, while other districts have notable gaps in permit information, particularly for general permits that were issued before the district's RAMS database system was fully operational. In addition, Corps staff are not notified of all activities authorized through its comprehensive general permitting authority, and may not enter some permits into the RAMS database.
How EWG Interpreted RAMS Nationwide Permitting Data
Each activity in a wetland or other water body that requires some form of action by Corps regulatory staff--a permit issued, denied, or withdrawn, a wetlands jurisdictional determination performed, or even a finding that no permit is required--is represented the RAMS system by a unique 'action ID' number. Each 'action ID' is assigned a special status code when some final regulatory determination is made by the Corps staff. Typically 'final status' codes indicate that a permit has been issued, denied or withdrawn. Just as a single housing project or road may affect several different streams or wetlands, one 'action ID' in the RAMS system may have several 'final status' codes, indicating that more than one permit was issued for a single project.
For the purpose of analysis, EWG examined all instances between January 1988 and June 1996 for which an 'action ID' in the RAMS database entered a 'final status,' indicating that a nationwide permit had been issued. EWG considered each of these 'final status' codes to correspond to one activity that was authorized by a nationwide permit. In all, nearly 137,000 final nationwide permit authorizations were represented in the in the database for the 27 Corps districts using RAMS.
For most nationwide permits, the RAMS system indicated that the activity was authorized under a specific nationwide permit. However, for some 'action IDs' no nationwide permit authorization was indicated. EWG counted these as nationwide permits of 'unknown' type.
For some 'action IDs' more than one nationwide permit was indicated. An examination of the notes accompanying some of these permits revealed that multiple associated with a single 'action ID' often indicated that a single activity was covered under two or more nationwide authorizations (for instance, one utility project might be covered both by the nationwide permit for utility backfill and by the permit for minor discharges). However, EWG found some cases in which several distinct activities were grouped under a single 'action ID'--for example, a development project involving several distinct utility lines, road crossings, and minor discharges into wetlands, all grouped together under a single 'action ID.' In such cases, EWG assumed that the permits in RAMS represented several separate uses of the different nationwide permits, rather than several similar activities all covered by several different nationwide permits.
Because a single action could be authorized by more than one nationwide permit, EWG's tally of the number of wetlands permits authorized in a given region was sometimes less than the sum of the number of actions permitted under the various nationwide permits.
Estimating the acreage affected by nationwide permits
Estimating acreage of wetlands and other waters affected by nationwide permits posed particular difficulties. The majority of nationwide permit actions contained no record of the number of acres affected by the permit. EWG is aware of not other source of information that would provide insight about the acreage impacts of Section 404 permits. Of the nearly 137,000 final nationwide permits in its permitting database, EWG found only 23,749 permits with some non-zero entry for acreage. Information on the acreage of wetlands mitigated was even more sparse. Furthermore, the average acreage affected by nationwide permits varied by year; varied significantly by state and Corps district, and varied even more significantly by the type of nationwide permit.
Although the acreage information in RAMS was limited, EWG did use the available data to estimate the actual acreage of wetlands and other waters affected by each permit in each region. Estimates of the acreage of wetlands or other waters affected by a particular nationwide permit in a state were conducted as follows:
If there was a significant number of permits of a particular type in a particular county in a particular year, with valid acreage estimates from the Corps, (e.g., more than 50 issuances of a particular nationwide permit in a particular county in 1995 with valid acreage data) , EWG assumed that all permits of that type, in that region, in that year, had the same average acreage as the permits in that year for which acreage data were available.
If there was not a significant number of permits of a particular type in a particular county in a particular year, EWG determined whether there was a significant number of permits of that type in that county from 1988 through 1996. If so, EWG assumed that all permits of that type, in that county, in that year, had the same average acreage as the permits of that type from 1988 through 1996.
If there was not a significant number of permits of a particular type in that county from 1988 through 1996, EWG determined whether there was a significant number of permits of that type in the state from 1988 through 1996. If so, EWG assumed that all permits of that type, in that county, in that year, had the same average acreage as the permits of that type from 1988 through 1996 in the state.
If there was not a significant number of permits in the state from 1988 through 1996, EWG assumed that permits with missing acreage data had the same impact as the national average for that permit. When several different nationwide permit numbers were associated with a single 'action ID', EWG apportioned the acreage among the various permits indicated.
Because of data inadequacies referenced above, EWG's estimates are subject to substantial uncertainties and significant sources of potential error:
Possible sources of underestimation:
- Permitting data missing from the RAMS system. According to the data currently in the Corps' RAMS system, fewer than 7,000 activities were authorized under the nationwide permit program in 1988, but 29,000 activities were authorized in 1995. Although the use of nationwide permits may have increased somewhat from 1988 to 1995, the bulk of this apparent increase is due to gaps in RAMS data in earlier years. Most districts began using RAMS in the early 1990's, and many of these districts did not enter a significant amount of permit data, particularly for general permits, for actions authorized before their RAMS systems went on-line. Because permit data for early years are incomplete, it is likely that EWG's wetlands permitting database significantly underestimates, perhaps by as much as one-half, the number of activities authorized by nationwide permits and the acreage affected by the program.
- Permits not entered into RAMS. Most Corps districts have made a concerted effort to enter new permits into RAMS. However, some Corps district staff admit that there are still some permitted activities that are not entered into the RAMS system (Note 3). Permits not entered into the RAMS system, either by accident or oversight, may result in undercounting of the number of activities and the acreage impacts that have taken place under the nationwide permit program.
- Activities of which the Corps is not aware. Corps staff can enter data into RAMS only when they know of an activity that is authorized under a permit. However, many nationwide permits give blanket authority for permittees to undertake projects in wetlands or other waters without notifying the Corps. NWP #26, for example, requires a permittee to notify the Corps only when an activity would affect more than one acre of jurisdictional waters; for activities one acre or less, no notification is required, and the activity can proceed without the knowledge of Corps staff. The Corps itself has estimated that more activities are authorized under nationwide permits without the Corps' knowledge as are authorized with the Corps' knowledge (Note 4). EWG's permit estimates derived from RAMS data may, therefore, very substantially understate the number and acreage of permits issued under the nationwide permit program, particularly for permits that do not require notification.
Permits without a specified location. For 2,900 of the 136,908 permits (2.1 percent) in the Corps' database, the precise county or state in which the permit was issued could not be determined. Data gaps in the location of permitted actions were highest in the Memphis district, where nearly 23 percent of nationwide permits in RAMS could not be identified in a particular county or state. Smaller, but still significant numbers of permits in the Kansas City district (7.5 percent), Mobile district (7.0 percent) and St. Louis district (6.8 percent) could not be located within a particular state. Although such data gaps do not affect EWG's totals for the number or acreage of permits issued by Corps districts, they may reduce totals for particular states or counties.
Possible sources of acreage overestimation:
- Small sample sizes. In many cases, EWG's estimates for the acreage affected by a particular permit in a county or state were based on a comparatively small number of valid acreages in RAMS. In general, small sample sizes increase the likelihood of error in making acreage estimates. For example, only 59 of the 1,428 activities (4 percent) authorized under NWP #3 (for maintenance of existing structures) in Maryland were associated with valid records of permit acreage. There is no way to tell from the RAMS data alone whether these 59 permits with valid acreages were a representative sampling of all 1,428 activities authorized under NWP #3. If Corps staff were more likely to record the acreage for activities with particularly large impacts, the limited sample sizes may have led EWG to overestimate the acreage affected by a given permit in a region.
- Permits of unknown type. For 23,486 activities authorized under nationwide permits, RAMS did not specify the number of the particular nationwide permit that covered the activity. Of these permits, only 373 (1.6 percent) have valid acreage information in the RAMS database. These 373 permits accounted for 328.4 acres of wetlands and other waters, or an average of nearly 0.9 acres each. In estimating the acreage impacts of permits of unknown type, EWG assumed that all permits of unknown type had the same average acreage as the unknown permits for which acreage was available. This may have led to an overestimation of the acreage of wetlands or other waters affected by permits of unknown type, particularly if Corps staff were more likely to record acreage information for permits affecting a large acreage.
- Permits affecting 'zero' acres. In some cases, final nationwide permits in the RAMS database are associated with data indicating that zero acres of wetlands or other waters were affected by a permit. It is unlikely that all of these permits affected zero acres--presumably, the Corps would not say that an activity is covered by a nationwide if it does not affect any of a wetland or other water body. It many of these cases, it is likely that Corps staff created a data record for acreage information in the RAMS database, but never actually entered the acreage data. In making acreage estimates, EWG treated 'zero acre' permits as if no acreage record existed for that activity. However, an alternate interpretation of 'zero acre' permits is that they affect a negligible or insignificant acreage of wetlands or other waters. To the extent that 'zero acre' permits represent activities with little or no acreage impacts, EWG may have overestimated the acreage affected by some nationwide permits.
It is important to note that impacts to wetlands or other waters, as represented in RAMS, do not necessarily correspond with losses of acreage. Some nationwide permits may authorize activities in wetlands or other water bodies that have little or no adverse impacts. For example, EWG estimates that 3,383 acres of wetlands or other waters in the 27 Corps districts examined were 'affected' by NWP #27 (for wetland and riparian restoration and creation activities). While some of these effects may have been harmful to some ecological functions and values (as when wetlands are 'restored' by being converted into deep water habitats), it is likely that wetlands functions and values for a significant portion of these acres were preserved or improved. In addition, compensatory mitigation may have been performed for some of the permits authorized under the nationwide permit program. EWG's totals for acreage impacts, therefore, should not be interpreted as indicating a net loss of acreage, but merely an estimate of the scope of the nationwide permit program's impacts on wetlands and other waters.