Like most rape investigations, the Tantra Lake serial rape case began with the terror and tears of a young woman who was being attacked by a man. It ended after years of complex and difficult work by the detectives of the Boulder Police Department.

The first victim was Sarah. On a cool September night in 1993, Sarah was sound asleep in the bedroom of her south Boulder condominium. She was awakened by a man holding a knife to her throat. His breath stunk of stale cigarette smoke. He warned that he would cut her if she didn’t submit. He then raped her.

Sarah’s attacker cut off her shirt, her pajama bottoms, and her panties. She was then naked, face down on her bed. He fondled her breasts, he kissed her, and then he attempted to vaginally penetrate her. He couldn’t maintain an erection, so he forced Sarah to turn over onto her back. She was forced to face him, as he mounted her and raped her.

The attack ended, and the rapist left Sarah alone in the darkness of her home. Like most rape victims she was shocked, disoriented, and still terribly afraid. She feared that her attacker would return and use his knife to cut more than her clothing if she tried to get help. But eventually, she summoned the courage to call 911.

Patrol officers from the Boulder PD responded. They searched for the rapist, but he was long gone. Inside her home, Sarah was interviewed by Boulder detectives. Although she was distraught, Sarah agreed to a full sexual assault examination.

On that night of Sept. 13, 1993, the detectives who interviewed Sarah believed that her testimony and the evidence taken from her body during the sexual assault examination would help them bring her rapist to justice. What they didn’t realize was how long it would take to close the case. The 1993 rape of Sarah was the first of five felony rapes that would take place in the large south Boulder condominium complex called Tantra Lake. And it would take more than a decade to catch the man who evidence pointed to as the “Tantra Lake Rapist.” Before he was captured, the same DNAand MO would be linked to eight rapes in three states.

IT’S THE SAME GUY

Five years after Sarah was raped, I became involved in the Tantra Lake serial rapist investigation. Much had happened in that half decade. Several more women had been raped in the Tantra Lake complex, and the Boulder PD had developed theories about the crimes.

We believed that we had a serial rapist on our hands. He clearly had a victim profile, showing a preference for single, petite, dark-haired women. In addition, the MO was the same in all of the attacks. He entered each victim’s home through a sliding glass door, immediately covered the victim’s face, and used a stocking to disguise his face. What he said to each victim was also similar. He assured his victims that he wasn’t going to hurt them, only look at them; then he made specific sexual comments and threats. And as Sarah had noted in 1993, he reeked of cigarette smoke.

But we had no way to confirm our theory. The Tantra Lake Rapist had been pretty careful about leaving behind evidence. He’d worn latex gloves and a stocking mask during the attacks. And the evidence that had been taken from the victims did not include enough DNA to run a conclusive test.

Then we got some help from technology. The technicians at the Colorado Bureau of Investigation Crime Laboratory updated the lab’s DNA capabilities to STR (short tandem repeat) analysis so that it would be compliant with the new FBI CODIS (Combined DNA Index System) database. The STR process permitted DNA matching of the evidence from the Tantra Lake rapes, and the conclusion was that all of the rapes were committed by the same guy.

Unfortunately, we didn’t know the identity of that guy.

INTERAGENCY COOPERATION

With each new attack, the pressure on us increased and more and more police resources were allocated to the investigation. At one point, we even rented an apartment and placed female officers as decoys in an attempt to lure the suspect. None of the tips, meetings, or interviews developed any solid suspects.

The case was leading nowhere. Then we got an unexpected break. DNA profiles from the Tantra Lake Rapist were uploaded into CODIS in 1998 by the CBI crime lab. And they got a hit.

DNA from the Boulder case matched the DNA left behind by a rapist during a 1994 attack in Lakewood, Colo., and a 1995 attack in Austin, Texas. Neither of these two cases had an identified suspect, but the DNA links were significant, especially the Austin case. It allowed us to make an educated guess that the suspect was a white male (based on victim reports) who had possibly lived in Boulder, Lakewood, and Austin. Using this theory as a basis of subsequent database searches, the suspect list was narrowed considerably.

The 1998 Boulder rape also led to another break in the case. Following this attack, a statewide law enforcement advisement was issued, asking other agencies if they had any similar, unsolved sexual assaults. And the Glendale (Colo.) Police Department responded, indicating that there had been a similar sexual assault in Glendale one month earlier. In addition, Glendale PD detectives told us about a January 1998 rape that had occurred in Denver, just a few blocks away from the Glendale address.

SEARCH ENGINE

As we waited for DNA test results on evidence resubmitted to CBI, we began using the Lakewood and Austin links to attempt to identify possible suspects. Again, the theory we used was that the suspect had lived in the three locations where we knew for sure he had committed rapes.

We started this search by going back through the original resident list given to us by the apartment management company, as well as the list of other possible suspects previously identified through tips or other means. We then queried these names through Accurint for Law Enforcement, an investigative product available to police departments and law enforcement investigators on a paid subscription basis.

Accurint is essentially a data clearinghouse, bringing together hundreds of public, government, and commercial databases into one user-friendly interface. Our experience showed that Accurint generally had address information going back around 15 years. What we were looking for, of course, was whether any of these individuals might have had previous addresses in the three cities where the rapes had occurred.

This query method proved to be a tedious process. We did find some individuals who fit our criteria, but they were all ultimately eliminated either through firm alibis or DNA comparison. All told, several hundred individual searches were done in this manner. Of this total, approximately 50 individuals were eliminated by DNA comparison.

OFF-LINE SEARCHES

Besides the Accurint queries, other attempts were made to identify our suspect by conducting off-line searches of the Texas Department of Public Safety Driver License database and the Colorado Department of Motor Vehicle (DMV) databases. Based on the theory that the suspect had a previous driver’s license from one of these two states, we requested a list from Texas of all male driver’s licenses where the person had a previous license from Colorado and requested from Colorado a list of all male drivers who had a previous license from Texas. To narrow the list, we requested that this information be limited to between approximately 1994 and 1996. We got tens of thousands of names.

But we didn’t let the size of the search discourage us. Using Excel spreadsheets to correlate and organize the data, we made some assumptions about our suspect’s height, weight, and age, and eliminated those individuals that didn’t fit within these parameters. We also narrowed the search by including only those individuals with an address within the general proximity of the victims’ addresses. We then used Excel’s “word search” function to find all individuals who had addresses in the same neighborhoods as our victims. This yielded some promising names. However, individuals identified using these methods were eliminated either by alibi or DNA.

In May 2004, I spoke with an Accurint representative about running an off-line search of the company’s enormous database. An off-line search is basically a manually entered query using predefined parameters. These queries can only be run by Accurint database specialists. Accurint’s database contains literally billions of pieces of data from hundreds of commercially available databases such as credit headers, department of motor vehicle records, and property ownership records.

After discussing the facts of the case with the Accurint programmer, we agreed on the following parameters for the query: All males with a date of birth in 1968 or after who had lived within a 10-mile radius of 80302 (Boulder) and had data records there between 1991 and 1999, and who had lived within a 10-mile radius of 80228 and had data records there between 1992 and 1996, and who had lived within a 10-mile radius of 78759 and had data records there between 1993 and 1997.

These parameters were based on what we believed was the youngest age the suspect could be and that the suspect was living in proximity to his victims during the time of the assaults. We used only the three cities where we knew for sure (based on the DNA links) that the rapist was definitely linked to. Those based on MO were not included. Additionally, the 10-mile radius was suggested by the programmer for the first query, reasoning that the area could be narrowed down if there were too many “hits.” As it turned out, this additional 10-mile “buffer” was crucial in identifying the suspect within this first query.

Within 12 hours of requesting this query, Accurint e-mailed us an Excel file that contained the results. There were 120 hits. After a couple of hours of sorting this database, one name, Bradford Wagner, stood out.

PRIME SUSPECT

Two things made Wagner look right for these rapes. First, he had a previous address at 1082 Dahlia Street in Glendale. This was significant because, although the Glendale assault is included in the list of assaults we believed the Tantra rapist was responsible for, it was not included as part of the search parameter since it was only linked by MO. Second, the data showed Wagner had a previous address at 5100 Leetsdale in Denver, half a block from the address of the January 1998 Denver rape.

The 10-mile buffer that Accurint suggested was crucial in identifying our suspect. Wagner had never lived in Lakewood. His name appeared on the list because the border of his Glendale ZIP code was 9.8 miles from Lakewood.

We got lucky. And in major crime investigations, a little luck never hurts. There were no sexual assaults in Wagner’s history, but there were some red flags. He had several burglary convictions dating back to the late ’80s, and had arrests in five different states during that time for offenses, including burglary, theft, and DUI.

We were becoming more and more confident that Wagner was our guy. We located two of his ex-girlfriends from the 1990s. Both women reported he was a heavy smoker and recalled that he liked to go for late-night walks alone. Both also reported being stalked by Wagner after their relationships ended. In one startling example, Wagner’s ex-girlfriend reported that she woke up during the middle of the night with the strange feeling that someone was in the bedroom with her. She was startled to find Wagner standing over her bed, wearing dark clothing. When she asked him how he got in, he nonchalantly told her he had climbed up to her second floor balcony and entered through her sliding glass door. Wagner told her he wanted to make sure she had come home alone. The woman did not report the incident until our interview. Wagner was looking real good for our rapist. We requested an NCIC off-line query from the FBI. This provided the date, time, and agency information of any query made regarding Wagner in the past 10 years.

The information showed numerous contacts, including a traffic stop in the Austin area during the same month that the Austin sexual assault had occurred as well as a contact in Lakewood. Lakewood detectives followed up on this information, and discovered that Wagner was contacted in 1994 with a woman identified as his girlfriend who lived in a complex within a couple of hundred yards from the 1994 Lakewood rape victim. The NCIC query also showed that Wagner had traveled numerous times outside the United States.

By this time in the investigation, we had also received results back from the CBI lab on the evidence we had resubmitted from our cold cases. Three other cases, two from 1993 and one from 1994, were found to have evidence that contained suspect DNA. The DNA profile matched the DNA from the other Tantra cases.

DNA SAMPLE

For the final link to Wagner, we needed to obtain a sample of his DNA. We believed Wagner was a flight risk because he had a passport and was known to travel extensively. The decision was made to collect his DNA without his knowledge.

We learned that Wagner was a licensed real estate agent working in Glenwood Springs, Colo., a mountain town close to Aspen. Posing as a potential homebuyer, Sgt. Pat Wyton contacted Wagner. Wyton set up a meeting with Wagner under the guise that he was looking to purchase a million-dollar home in the Glenwood Springs area.

Wagner and Wyton met for lunch on May 20, 2004. Undercover detectives from the Boulder County Drug Task Force were also on the scene providing support and surveillance. Wyton entered the restaurant and shook hands with Wagner. Wyton had planned on collecting the utensils and drinking glasses used by Wagner, but the waitress removed these items from the table before they could be collected. At the conclusion of the meeting, Wagner and Wyton shook hands for a second time. Upon exiting the restaurant Wyton’s right hand was swabbed with moistened sterile swabs in an attempt to collect any epithelial cells that might have sloughed off Wagner’s hand during the handshake. These swabs were submitted to the state forensic laboratory and provided the first DNA link between the Tantra Lake serial rapist and Bradford Wagner. Before leaving Glenwood Springs, undercover detectives collected swabs from the handlebars of the bicycle Wagner was seen riding, as well as swabs from his car door handle and the door knob to his residence.

On June 9, 2004, the CBI Crime Lab completed the initial DNA analysis on the samples acquired by Boulder PD officers in Glenwood Springs. The results confirmed the DNA link between Wagner and the Tantra Lake Rapist. Wagner was arrested at his real estate office on June 11, 2004, by members of the Boulder and Glenwood Springs police departments. He refused to make any statements to law enforcement. He was transported back to Boulder and lodged on the arrest warrant. Detectives also served a search warrant on Wagner’s residence and obtained buccal swabs from Wagner for confirmatory DNA testing.

The criminal case against Wagner was strong. At the preliminary hearing all counts of sexual assault were bound over into District Court. But just one day before the scheduled July 14, 2005, motions hearing, we received a call from the Boulder County Jail. Wagner, 37, was found dead in his cell from an apparent suicide. To some of the victims Wagner’s death was justice denied; they were upset and angry that he wouldn’t have to face them in court. Other victims felt that he had received the ultimate punishment. Either way, the Tantra Lake Serial Rape Case was closed.

Chuck Heidel is a 21-year veteran of the Boulder Police Department. He is currently assigned to the Detective Section Major Crimes Unit.