War Stories: Lessons Learned with Plate Frame Heat Exchangers

A building located in the downtown area of a large city is where our war story begins. A central plant serving the building and several other buildings throughout the downtown area had monitors that read GPM flows and temperature differences to calculate the BTU loads and bill the customer for their usage. The steam system monitored the pounds per hour of steam requirements to bill the customers, as well.

With a total of 30 floors, the building contained a single heat exchanger and pumps that circulated water for the different chilled water piping risers. The customer reached out, stating that the current plate frame heat exchanger could not hold the facility temperatures at setpoint with the chilled water supply provided. The building had been experiencing problems with air handling units (AHUs) on the floors not holding the 55oF discharge air setpoint per the original mechanical design.

Upon my initial visit to the site, I reviewed the design documents to get an idea of the original design intent. Figure A illustrates the arrangement of the piping and heat exchanger orientation.

The facility manager asked me to review why the building could not maintain proper conditions for the tenants, resulting in areas being warm and humid. We reviewed the cut-sheets for equipment.

The plate frame heat exchanger has a certain approach or water difference for heat transfer. Typically, the required incoming inlet water from the central plant is 39°F to 42°F (sometimes referred to as cold-side) and the exiting side of the plate frame heat exchanger for the building is 42°F to 45°F (referred to as the hot-side). We refer to this as a 3°F approach.

Through past observations, I knew plate frame heat exchangers could get contaminated or contain film buildup on the plates, thereby reducing the heat transfer. This not only causes the system to be unable to perform as designed but also reduces efficiency. We typically see the differential pressures exceed the original design pressures. The higher differential pressure can be misleading at times, assuming that the GPM flow is higher and not lesser flow. In this case, GPM flow stations showed otherwise. I made a point to take pictures showing before and after results.

As we can see with Example 1, the system indicated a flow of 380 GPM, and the entering water temperature was 42.8°F, while leaving temperature was 53.6°F. Notice the calculated tonnage of 171.8 tons. When we do the math, (the system has no glycol and is 100% water) 500 x 380.3 GPM x (53.6 – 42.8) delta T = 2,053,620 BTUH/12,000 (BTU per ton) = 171 Tons. The meter showed 171.8 tons which is really close to the hydronic BTU equation.

On the other side, the leaving side of the plate frame heat exchanger to the building is at 49°F. This places the approach to as much as 6°F, helping me to determine that the plate frame heat exchanger was dirty and contaminated. Another good indicator was that the plant side of the system had pumps that would energize and modulate the VFD speed to accommodate the proper differential setpoint and maintain the 12°F delta temperature. The central plant’s sequence of operation says water returning to the plant cannot be colder than 53°F. The booster pump VFDs would increase and decrease speeds to maintain the water on the plant side to be at 53°F. As Example 1 shows, the temperature was already at 53.6°F which had both booster pumps at 100% VFD speeds, yet still did not satisfy the tenant requirements.

The plate frame heat exchanger had over 746 plates. Trying to break down the plates and attempt to clean the plate frame heat exchanger would require a lot of labor hours and downtime for the facility, which they could not afford. The heat exchanger was also over 30 years old, so new gaskets for the assembly were not readily available.

I expressed concerns to the facility manager that I was not comfortable separating the heat exchanger for cleaning. Due to the age of the equipment and possibly not having been exercised for a great deal of time, I also mentioned concerns that the isolation valves for both the plant side and the building side would not isolate and hold.

At this point, we recommended chemical cleaning. I made sure that the customer understood I was not sure how well this would work, but we would do our best. We subcontracted a chemical cleaning contractor to do the cleaning and provided the connection points for the portable pump assembly to circulate the exchanger with the necessary chemicals. Pipefitters exercised the isolation valves, drained the water off the plate frame assembly, and installed the hose connections for the chemical cleaning team. Example 2 shows two pictures of the chemicals utilizing catch basins to circulate the chemicals while ventilating the room to maintain a safe environment.

Per the chemical treatment company, we discovered that the heat exchanger was extremely contaminated with a bacteria type organism. The secondary side water quality was good, which had the proper chemical inhibitors.

Once the plate frame heat exchanger was cleaned, we flushed the heat exchanger thoroughly with fresh water to confirm the exchanger was chemical-free by taking water samples of water that was drained from the system. At this point, we filled with water and vented the heat exchanger to confirm the plates had fluid back in the system. The primary concern was that if the main isolation valves were opened to fill the plate frame heat exchanger, the inrush of water could cause issues with gaskets/seals and cause the heat exchanger to start leaking.

As the plate frame heat exchanger was filled, the main isolation valves were opened very slowly to allow water static head of the building and the plant loop pressures to come into the device at a very slow rate. Once everything was stabilized with no leaks, we placed the pumps back online for both the building side and central plant side. We had the building down for over 12 hours before returning the building back into an occupied state.

With all equipment operational, we saw the entering water was 39.6°F and the leaving water was 52.8°F, which was less than the 53°F requirement from the central plant as the booster pumps remained off. We also saw the flow rate for the central plant side increase from the previous 380 GPM to 1526 GPM and the overall tonnage went from 171 to 842 tons. The heat exchanger was restricted with little to no heat transfer until cleaned. The approach was still near 6°F, but the actual chilled water setpoint of 53°F had the return value raised to 48.6°F.

With the cleaning of the plate frame heat exchanger, we accomplished the task at hand and made another customer very happy. The facility manager was very grateful that the plate frame heat exchanger was performing considerably better, and the occupants were grateful as well.

I feel blessed to not only have the ability to apply the theory, but also apply the practical aspect of things as well from being a NEBB Certified Professional. There is a certain amount of gratitude and accomplishment doing this work while applying TAB formulas. We are supporting our local communities with the information that NEBB offers in our TAB Procedural Standards. There are certain things that books can help teach us, but being in the field and applying the technical side of things is critical in troubleshooting and continuing to grow and learn every day.