Windscreens improve performance, reduce O&M cost of air-cooled condensers

Air-cooled condensers depend on steady air flow created by a properly designed system of axial-flow fans, normally elevated significantly above grade. Induced by the fans, ambient air then flows vertically through the tube bundles above, condensing the steam within those tubes.

During high winds, the condenser outer shell (wind wall) deflects the ambient air, producing a jet stream below, along the fan inlet region. Jet stream conditions lead to less air flow (suction starvation), reduced pressure, mechanical stress on the fan blades and gear reducers, and increased backpressure on the system, reducing steam-turbine output. An extreme crosswind can mean greatly reduced air uptake, fan stalls, blade damage, and costly motor and gear/drive maintenance and repair.

Although such conditions are both location and time specific, some stations have turned to a system of windscreens (shields) to eliminate, or at least reduce, this jet stream effect and minimize unfavorable air flow patterns,

Overall ACC performance. Wind effects vary dramatically from plant to plant, and from season to season. Factors also include ACC orientation onsite, the design sizing of the ACC and fans, and the design of the steam turbine. Nearby buildings, storage tanks, industrial facilities, and trees also can have an impact. There is no one-screen-fits-all solution.

If wind is a maintenance or performance factor, the common goal is to design and place screening around the perimeter of the fan system to create uniform air flow into the fans, reduce vibration and stress created by variable winds, and normalize the amperage on the motors driving the fans. Screens in a cruciform pattern under the ACC will also help enhance air flow into the fans. Balanced and efficient fan operation should then limit equipment damage and improve thermal performance. Service and maintenance should be more predictable.

To date, most windscreen installations are retrofit projects. Many benefits can be measured, but others are still in review. The common parameters studied are:

      • Fan performance.

      • Blade maintenance.

      • Gearbox and motor damage.

      • ACC thermal performance.

      • System backpressure.

      • Steam-turbine output.

Perhaps the best summary comes from the 2015 ACC Users Group meeting in Gettysburg, and the core information is included and updated here. Engineers know that wind effects on ACCs is drawing increased global attention, and that impacts include plant thermal performance, fan blade damage, and cell-by-cell fan duty, among others. It is difficult, however, to quantify and measure all thermal-performance issues, backpressures, cell-by-cell performance, and other specifics.

But with more than 50 ACC units worldwide already using windscreens or shields, wind-effect experience and knowledge are growing quickly through commercial case studies, smoke tests, CFD analyses, and a variety of field and wind tunnel examinations.

Materials and options. Screen and shield materials vary, including local resources. Most installations, however, use a fabric system in a permeable mesh format ranging from about 40 to 75% solid, and with a range of pressure drop coefficients. Screen placement configurations vary; the outer perimeter location is the most common, elevated around the fan inlet section of the ACC structure on site-specific sides. Such installations must often allow for structural bracing, pipe work, cable routings, and other interferences, especially for retrofit projects.

Early British tests, results. ACC windscreen installations began in 1998, in the UK at the 360-MW King’s Lynn combined-cycle station in Norfolk, now owned by Centrica Energy. This installation, by Galebreaker Group, is cruciform (cross-shaped) with the screens going from grade up to the fan housing. Of similar design, was the first US installation in 2003 at Reliant Energy’s Bighorn Station (now NV Energy’s 530-MW Walter Higgins Generating Station) in Primm, Nev, about 40 miles south of Las Vegas.

Early studies in the UK determined that high winds under ACCs increased system backpressure and thereby reduced power output. King’s Lynn, commissioned in 1997, decided the following year to install the cruciform screens, from the underside of the fan units to ground, and selected PVC-coated polyester mesh, 55% solid.

Operators found that with these screens, ACC vacuum improved when ambient winds increased, because of improved fan performance. Measured in 2006 with average winds of 9.6 mph, vacuum had improved an average of 0.165 to 0.178 in. Hg, boosting overall performance. In 2011, perimeter screens were also added (with one side motorized).

Also in the UK, Coryton Power Station is a 753-MW combined cycle commissioned in 2002. Operators determined that high winds under the ACC increased system backpressure and therefore reduced power output. Further research at Coryton included CFD modeling. A major concern beyond wind shear under the fans was the low-vacuum impact on the LP turbine.

The station added Galebreaker perimeter screens at various heights in 2004. Screen material was identical to that used at King’s Lynn. In 2005, based on average wind speed of 8 mph, vacuum had improved an average 0.148 in. Hg.

Caithness Long Island Energy Center. Perhaps the most detailed and long-term windscreen study is at the 350-MW Caithness Long Island Energy Center, an ACC windscreen retrofit project and analysis that began in 2012. Caithness was first placed into operation in 2009.

Caithness, in Yaphank, NY, is a 1 × 1 configuration with an 18-cell ACC (three streets with six fans each). During the first period of operation the ACC experienced cracking of fan blades and vibration issues, as well as a few motor trip problems.

The aerodynamic impact of wind on fan operations was then studied. Early ACC testing, including fan intake smoke analysis, was a joint effort by Caithness, Siemens Energy, and Howden Group Ltd. GEA Group supported with structural input and reviews.

The site is subject to high winds. According to Bill Wareham, site manager for Siemens, “We are five miles from the south shore of Long Island, with prevailing winds out of the southwest off the ocean. Wind speeds of 10 to 20 mph are fairly common, and we have had two major storms over the past five years (Hurricanes Gloria and Sandy). On a few occasions each year we get wind speeds approaching 50 mph.”

Various screen configurations were studied and Caithness was outfitted with retractable perimeter windscreens by Galebreaker, a unique rolling feature at time of installation in 2012. The screens can withstand 120-mph winds by design. Also during this retrofit the six-blade fans were replaced with nine-blade fans to address vibration and loading concerns.

Retractable screens were selected because of the potential hurricane-force winds. With fixed screens in place, such winds could exceed the structural limit of the ACC.

Then, because the screens were retractable, the site was selected for in-depth study, and for determining the effects of screens in various deployment conditions. When fully deployed, the screens cover approximately half of the grade-to-deck vertical dimension.

During the next three years, comprehensive research generated a vast amount of data with selected measurable benefits, primarily more uniform inlet velocity and a significant reduction in dynamic fan-blade loading.

Study participants were Maulbetsch Consulting, Galebreaker (windscreens), Howden (fans), Senta Engineering LLC (CFD), University of California-Davis (wind tunnel), and Caithness (host site and test operations). The study was funded by the California Energy Commission. A few testing specifics follow.

The immediate impact of adding screens seemed to be improved (more uniform) air flow entering the windward side perimeter fans and reduced stress on fan blades.

Fan data were collected for 18 months. For a period of two months in 2014, data were gathered for two cells (3.4 and 2.4) at full load and with all fans at full power. Fan inlet velocity was then plotted against wind speed measured at the storage tank wind vane. Data sorting parameters included wind direction and screen position (0%, 25%, 50%, 75% and 100% deployment).

For the two cells combined, at average air inlet velocity:

      • Screens had a positive effect (fan flow rate) at higher wind speeds.

      • Screens had a negative effect (fan flow rate) at lower wind speeds.

      • A windspeed of 9 mph was breakeven. Screen deployment offered the best results (up to a wind velocity of 18 mph).

Data also were gathered for:

      • ACC recirculation (heated plume air back into the inlet stream).

      • Steam-turbine backpressure.

      • Dynamic fan-blade loading.

For blade dynamics, screen deployment showed significant benefit. Screens reduced dynamic blade load at higher wind speeds by a factor of two to three. Thermal performance and backpressure benefits were not as clear.

A physical wind-tunnel model was created at UC-Davis to advance both depth and breadth of the wind and windscreen studies at Caithness. Study results showed positive correspondence with field data, an accurate physical representation for clarity and understanding, and the ability to explore alternatives. In this case, alternatives were:

      • No screens.

      • Perimeter screens.

      • Perimeter plus cruciform screens.

CFD modeling was created that gave highly detailed representations of the ACC and its surroundings. These models qualitatively showed the upstream complexity of the inlet boundary layer, upstream obstructions (trees and buildings), and the impact of adding a horizontal ledge at the bottom of the ACC wind wall. Quantitative results under windy conditions were not achieved. However, air velocity entering the fans is now more uniform with the screens, and dynamic blade loading has been significantly reduced.

Wareham points out that normal operation is now with windscreens 50% deployed.

Posted in 7F Users Group |

Comments are closed.