Narrative:

I recognized aircraft X and aircraft Y were going to be an issue at 38;000 ft. I took a few minutes to analyze all of the possible solutions for the vector I was going to use. There was another aircraft at 38;000 ft that would eventually be an issue with aircraft X; so I decided to vector aircraft X behind aircraft Y. It was a shallow angle; and I knew that I was taking someone who was slightly behind and keeping him behind; requiring a good deal of turns. I turned both aircraft out to the right. At the point of closest proximity I measured at 4 minutes away; 3 measurements; 5.2 miles; 4.9 miles; and 5.4 miles (to my recollection). Since I had one measurement under 5; I turned the aircraft more. At 2 minutes; I measure 4 times; 5.2; 5.2; 5.0; and 5.2. Since all measurements were 5.0 or greater; I felt comfortable that I had run a safe and efficient vector between the aircraft. Under the new eram system; conflict alert is programmed to flash at 5.5 miles or less. Since this vector was going to be less than 5.5 miles; conflict alert was activated (nuisance alert). I felt comfortable with the mileage I measured. At 1 minute away; I got a measure of 5.2; and still felt comfortable about what I was doing. I could see in the eram conflict alert view that I had plenty of mileage between the aircraft. At one point; I looked to the conflict alert view and saw the eram system said the aircraft were 5.0 miles apart. This happened on 2 hits (0.2 miles closer than I estimated); on the next hit; the mileage showed 4.99 (0.01 under 5 miles - and therefore I am convinced of a loss of separation.) the following hit showed 5.11 miles; and then the conflict alert stopped. Conflict alert in host was set to 4.9 miles. It gave you an alert when the computer believed you weren't going to have separation. It was not 100% accurate; but nonetheless; consistent with loss of separation. Having the conflict alert go off at 5.5 miles; causes one to ignore the alert when running a tight vector. I have run several vectors since eram was implemented at 5.4 or 5.3 miles apart. Seeing the nuisance alert when you have separation allows the controller to begin to disregard the alert; and sadly to a point where the controller should probably do more; as in the case above. I think it is a very poor design; and should be changed to flash only when it is believed that separation will be lost; not the when separation exists but will be close.

Related reports:

Google
 

Original NASA ASRS Text

Title: ZAB Controller voiced concern regarding the new software parameters of the ERAM Conflict Alerts equipment; noting that the HOST system was more preferable.

Narrative: I recognized Aircraft X and Aircraft Y were going to be an issue at 38;000 FT. I took a few minutes to analyze all of the possible solutions for the vector I was going to use. There was another aircraft at 38;000 FT that would eventually be an issue with Aircraft X; so I decided to vector Aircraft X behind Aircraft Y. It was a shallow angle; and I knew that I was taking someone who was slightly behind and keeping him behind; requiring a good deal of turns. I turned both aircraft out to the right. At the point of closest proximity I measured at 4 minutes away; 3 measurements; 5.2 miles; 4.9 miles; and 5.4 miles (to my recollection). Since I had one measurement under 5; I turned the aircraft more. At 2 minutes; I measure 4 times; 5.2; 5.2; 5.0; and 5.2. Since all measurements were 5.0 or greater; I felt comfortable that I had run a safe and efficient vector between the aircraft. Under the new ERAM system; conflict alert is programmed to flash at 5.5 miles or less. Since this vector was going to be less than 5.5 miles; conflict alert was activated (nuisance alert). I felt comfortable with the mileage I measured. At 1 minute away; I got a measure of 5.2; and still felt comfortable about what I was doing. I could see in the ERAM conflict alert view that I had plenty of mileage between the aircraft. At one point; I looked to the conflict alert view and saw the ERAM system said the aircraft were 5.0 miles apart. This happened on 2 hits (0.2 miles closer than I estimated); on the next hit; the mileage showed 4.99 (0.01 under 5 miles - and therefore I am convinced of a loss of separation.) The following hit showed 5.11 miles; and then the conflict alert stopped. Conflict alert in HOST was set to 4.9 miles. It gave you an alert when the computer believed you weren't going to have separation. It was not 100% accurate; but nonetheless; consistent with loss of separation. Having the Conflict Alert go off at 5.5 miles; causes one to ignore the alert when running a tight vector. I have run several vectors since ERAM was implemented at 5.4 or 5.3 miles apart. Seeing the nuisance alert when you have separation allows the controller to begin to disregard the alert; and sadly to a point where the controller should probably do more; as in the case above. I think it is a very poor design; and should be changed to flash only when it is believed that separation will be lost; not the when separation exists but will be close.

Data retrieved from NASA's ASRS site as of July 2013 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.