Recognized internationally for a name and even a suffix that has entered the pop culture lexicon, the luxury complex of modernist buildings on the banks of the Potomac River known as the Watergate Complex is an icon like no other in Washington. As a pioneering example of urban redevelopment, the compound is significant both for its architecture and planning. The design of the complex was conceived in 1961 and was substantially complete by 1971. The seven component buildings, interconnected by underground garages, are a combination of co-op apartments, hotel, retail, and office buildings. With its expansive views over the Potomac River, the Watergate is a mix of uses creating a self-contained and self-sufficient community.

The Watergate office building, located at 2600 Virginia Ave., is recognized for its notorious position in American history as the location of the bungled break-in at the Democratic National Committee headquarters during the presidential campaign of 1972. The Watergate Office Building, a 210,000-square-foot, 11-story office building built in 1966-1967, was in need of an HVAC perimeter system tuneup in 2019. Over the years, it has been adding chronic operation and maintenance problems to its dubious past reputation, so the property management firm, Joe Duffy, senior property/asset manager, Penzance, and his owner representative, David Avedesian, P.E., RPA, of Newport Associates, began to build a troubleshooting team to retro-commission the HVAC system that served the below-grade office space and 11-story office space above. The team would eventually include myself as well as the following:

  • Jonathan Eddy, lead engineer, Penzance, Watergate Office Building;
  • Joe Severt, senior project manager, S&W Controls Inc., Upper Marlboro, Maryland;
  • Mary Anne Kirgan, design engineer, owner, Systems 4 Inc., Gaithersburg, Maryland; and
  • Michael Stewart, application engineer, Siemens Controls, Washington, D.C.


In lieu of a traditional 2- or 4-pipe perimeter fan coil unit (FCU) system, this facility has a hybrid 3-pipe system. The cooling system is a classic 2-pipe chilled water supply (CHWS) and chilled water return (CHWR) system. The heating system is served by a monoflo system. This monoflo system is a one-pipe system connecting a FCU hot water supply (HWS) and return (HWR) to each heating coil. This monoflo system was a popular system in the 1960s. The building perimeter HVAC consisted of FCUs with a separate air-handling unit providing the ventilation directly to the office building. Each FCU had an automatic temperature control (ATC) valve on the CHWS along with a CHWR pipe connection. Each FCU also had an ATC valve on the single-pipe HWS from the monoflo main to the heating coil and its associated HWR back to the monoflo pipe main. 

Just like an automobile requiring routine annual maintenance and an eventual engine tuneup, this 3-pipe FCU system was in need of a retro-commissioning tuneup. Over the years, the facility staff addressed primarily chronic heating problems and less frequent cooling issues leaving the occupants unhappy on a regular basis. Attempts were made to satisfy the occupants with minimum success.

Issues with this existing 11-story, 3-pipe FCU system included excessive air entrapped in the FCUs at the 11th floor and excessive city water makeup to the heating side of the system.

Initial concerns raised with this hybrid 3-pipe system included: 

  • The possibility of hot water and chilled water being cross-connected because of the amount of air entrainment;
  • Inadequate heating requiring a booster pump; and
  • The ATC operation was in question, so the system operated in a semi-automatic control sequence with certain adjustments made manually on an as-needed basis.

Overall, this 3-pipe perimeter heating and cooling system was not operating based on the original design intent, and this was complicated with several corrective measure efforts over the years. So, it was time to step back and apply these traditional quality control steps versus continuing to jump to a solution that included four phases: data collection, data analysis, solution planning, and solution implementation. 



With this troubleshooting team, data collection included taking temperature and pressure gage readings at the pump; at each floor branch runout to the floor; and at the highest point in the system, which was the 11th floor. The team used only one gage to record the readings to eliminate the potential for multiple gages not being calibrated exactly to each other.

The first readings were taken at the hot water pump in the B3 Level basement floor with pump inlet and discharge pressure readings while in operation and also with the pump deadheaded (discharge pipe shutoff valve closed). Reviewing the pressure readings and, in particular, the deadhead gage reading, it appeared the pump curve information did not match up. Later, in the solution planning phase, this pump was taken apart. We found that the pump impeller was a ½-inch diameter larger than the shop drawing submittal pump curve. A new pump curve was created based on the actual pump impeller to identify the actual gpm based on the actual pump head. In the solution planning phase, a sonic waterflow meter was installed in the HWS main after the pump discharge to compare the gpm flow to the revised pump curve as well as confirm there was adequate HWS flow for the entire system as originally designed.

With the benefit of a hot water heating system riser diagram sketch, the team set off to take HWS and HWR pressure readings at each floor, working their way up the building while subtracting the existing static pressure at each floor elevation created by the column of water inside the pipe. This effort would be sufficient to create a hydraulic model of the HWS and HWR system. Several times, the readings taken did not make analytical sense. With further data analysis, it was concluded that sediment in the pipe system was affecting the pressure readings. The data analysis phase showed that the HWS to HWR riser diagram data collection of pressure readings at each floor indicated that no branch runout to any floor, and in particular the problematic 11th floor, needed a booster pump to overcome perceived inadequate pumping capacity. 

Using the same pipe riser diagram, HWS and HWR temperature readings were taken at each floor using an infrared heat gun. This approach proved to be marginally successful, with many of the temperature readings questionable. The solution planning was to confirm the HWS and HWR temperature readings at the steam-to-hot water heat exchanger in the B3 Level and at the top of the 11th floor. With this confirmation, the HWS temperature readings were adequate throughout the system.

An essential analytical tool for problem-solving an HVAC system was the team’s use of a freehand system riser diagram sketch (see Figure 1) to document temperature and pressure readings. “A picture is worth a thousand words,” and this sketch proved to be essential in supporting the team’s findings leading to eventual solutions. 

When creating the riser diagram, it was noted that, at the highest point in the system, the HWR riser had an automatic air separator where the HWR flows down through the building to the B3 Level basement equipment room hot water heat exchanger. This system flow diagram brought out the fact that the HWS/HWR monoflo main serving the 11th-floor FCUs was located at the underside of the 11th floor and that HWS would rise up into an 11th-floor FCU before returning back down to the monoflo main. This created individual high points at each 11th-floor FCU. The solution planning of the flow diagram showed that, in addition to the automatic air vent at the top of the HWR riser, there was a need to install individual automatic air vents at each of the 44 perimeter FCUs on the 11th floor (see Figure 2). 

When creating the system flow diagram, it was noticed that a system pressure relief valve had been installed at the inlet to the monoflo hot water heating pump. Each day, the relief would let go and send hot water through a relief pipe to a floor drain. Per the hot water heat exchanger (HX) manufacturer’s installation diagram, this relief valve should have been at the outlet from this HX and not installed between the air separator and the pump inlet. This finding led the inspection team to the system’s expansion tank, which proved to be undersized. The solution planning determined that the relief valve would be relocated per the manufacturer’s recommendation. That expanded tank size resulted in a larger tank. These two changes would eliminate the pressure relief from going off and resolve the excessive water makeup volume that had been tracked daily by the team.

During this data collection phase, it was determined that, although the HWS and CHWS system pumps had variable-speed drives (VSD), the differential pressure (DP) head control transmitter controlling each VSD was located across the associated pump inlet and discharge pipe. These DP devices should have been located on the CHWS, CHWR, HWS, and HWR lines at the furthest point in each system to maximize the VSD performance and energy savings. The solution planning phase directed us to relocate the DPs to the branch runouts serving the 11th-floor FCUs in advance of the retro-commissioning process starting.



With the solution planning phase complete and implemented, the retro-commissioning functional performance test (FPT) demonstration was created. Overseeing the building automation system (BAS) operator and following the BuildingSmartSoftware FPT directive, the BAS operator would drive the HWS and CHWS automatic control sequences of operation for the FCU perimeter system through the retro-commissioning FPTs with the commissioning engineer looking over the technician’s shoulder. At the same time, the S&W controls project manager verify the action-reaction of the FPT demonstration.

With each sequence demonstrated, e.g., “outside air temperature at 50°F or less” sequence, the ATC technician would print out the points and device results along with the system’s graphic for this sequence (see Figure 3). The commissioning engineer would collect each sequence’s performance together with the written FPT narrative using the corrective action log when necessary. This information would be later included in the commissioning report. Photographs of certain deficiencies would also be included in the report.

As a result of this retro-commissioning initiative, the team found numerous FPT deficiencies: The relocated DP transmitters serving the heating and cooling VSD pumps malfunctioned, and the sequence of operation could not be demonstrated; it was determined that both VSDs (heating and cooling) would not vary the speed as expected and were in need of repair; the steam-to-hot water heat exchanger, which was designed to module one-third and two-third steam control valves in series, actually operated in parallel; the lead HWS pump did not start when automatically signaled on and the standby/lag pump started; FCUs on the 11th floor that were supposed to shut off in the unoccupied mode of operation continued to run; and the equipment labeled on the ATC computer graphics did not match up to the equipment label in the equipment room (e.g., P-6 on the graphic screen was labeled P-1 in the equipment room).

On a more positive note
The flow meter confirmed the revised pump curve gpm and that there was adequate gpm, HWS temperature, and system pressure throughout the system, eliminating the need to add a booster pump to serve the 11th floor; the city water cross-connection issue was resolved, with the problem being the relief valve location and an undersized expansion tank; the automatic air vents at each FCU on the 11th floor resolved these terminal units becoming air-bound on a regular basis, saving endless hours of building engineer’s labor hours; and once the retro-commissioning corrective action log issues were completed, the building’s ATC system was able to operate on automatic control, eliminating any manual control requirements.


Retro-commissioning and troubleshooting problematic HVAC systems require a quality control approach to the solution plans. All too often, individuals will want to jump to the solution based on their experience, and while their solution may be correct and/or improve the situation, it takes data collection and analysis before implementing a solution.

The approach taken at the Watergate Office Building in 2019 took teamwork. Each team member brought a wealth of experience. From there, the HVAC system’s issues and concerns were revisited collectively. Through the patient effort of following a step-by-step process, the proposed solutions were implemented, and, once additional corrective action items were identified, success was achieved. 

The benefits to the above include no more temperature calls from the tenants; no more building engineering labor being spent on venting 44 air-bound FCUs; better temperature control has resulted in lower energy costs; better FCU control has resulted in lower electric costs; better control of the system pumping and HX has resulted in more reliable operation and less makeup water; no wasted labor hours by the building engineer responding to “I’m too hot/cold” work tickets; and tenant satisfaction for temperature, humidity, and ventilation control, and comfort is at an all-time high.