Adnan Alsaffar (Member,
IAHR)
Senior Advisor in Hydraulic and Marine Engineering
Bechtel Corporation, 5275 Westview Drive
Frederick, Maryland 21703-8306, USA
Ibrahim El-Desouki
Deputy Director
Hydraulic Research Institute
Delta Barrage, Egypt
Eng.
Abdel Meguid Radwan
Vice Chairman for Projects
Egyptian Electricity Authority
Cairo, Egypt
Abstract: The Ayoun Moussa thermal power plant is located on the Gulf of Suez in Egypt. The plant’s two gas/oil-fired units, which generate a combined 640 MWe, use a once-through cooling water system. Water is withdrawn from an offshore open channel with earth-filled embankments and returned to the sea through two offshore buried pipes. These pipes terminate at a box-type structure discharging at the water surface through a rectangular opening in the wall of the box. The outfall is located in a shallow water depth in proximity to the Suez Canal. This raised the concern that scouring of the gulf bottom by the surface dis-charge jet could lead to sediment deposition in the canal. Another issue is the potential for warm water recirculation into the offshore channel, whose inlet is located approximately 1,500 m from the outfall. Because of the complex nature of the water source and the afore-mentioned requirements, both analytical and thermal hydraulic physical model studies were employed.
Keywords: power plant, offshore intake, skimmer wall, offshore discharge, distorted thermal physical model, erosion protection, heat transfer, riprap sizing
In 1994 the Egyptian Electricity Authority (EEA) awarded Power Generation Engineering Services Company (PGESCo) the contract to design, procure, and construct the Ayoun Moussa thermal power plant. The plant is located on the Eastern Shore of the Gulf of Suez about 3 km south of the Port of Suez, Egypt. The plant consists of two 320 MWe dual-fuel units with a once-through cooling water system. The cooling water discharge is approxi-mately 26 m3/s, with a temperature rise across the condenser of 8.5 ºC. The overall layout is shown on Figure 1. Key considerations in developing the cooling water intake and discharge are:
l Locating the intake and discharge structures in shallow water
l The presence of the Suez Canal near the intake and discharge
l The potential for warm water recirculation into the intake
Fig. 1 General location map
This paper describes the field, analytical, and physical thermal model studies performed to arrive at a technically and economically feasible design.
To address the key issues, the following studies were performed:
(1) Detailed hydrographic survey encompassing an area with a maximum length of approximately 8 km and a maximum width of 4 km
(2) Development of a conceptual design to (a) select an intake and outfall that minimize the impact of siltation on the Suez Canal and (b) minimize the impact of construction and operation on marine traffic in the region
(3) Analytical studies to estimate (a) degree of warm water recirculation into the intake and to select details of the intake and (b) extent of erosion caused by the discharge and development of protective measures to mitigate local erosion
(4) Physical thermal model studies for the near field and the far field to define the extent of the erosion and to determine the degree of warm water recirculation into the intake, respectively
The bathymetric survey covered an area of approximately 26 km2 extending approximately 4 km offshore. The survey included current, temperature, bed sediment sampling and analysis, and wind speed. The current in the area of the intake and outfall during the survey ranged from 0.04 to 0.12 m/s, with an average of 0.08 m/s. The maximum current measured in the survey area was 0.20 m/s. Tidal level variation in the region from the survey performed during January 1994 to September 1995 had a mean range of 1.75 m and a maximum range of 2.20 m. Bed sediment samples collected at four stations showed D15 = 0.14 mm, D50 = 0.18 mm, and D85 = 0.50 mm. The bathymetry ranged from –0.50 m near the shoreline to –1.0 m at 750 m and reaching –3.5 m in approximately 1,000 m.
A conceptual design considered the power plant layout, the owner’s preferences, and the presence of transmission lines. Various options were evaluated, including offshore intake with a velocity cap, and shoreline discharge. The selected alternative consists of an offshore intake channel with earth-filled dikes and an offshore discharge through two buried pipes connecting to a discharge structure. The intake consists of an offshore dredged channel with protective earth-filled dikes and skimmer wall located at the inlet to the onshore intake pipes. The intake channel with invert at –5.5 m extends offshore approximately 1,600 m to a location with seabed at –5.75 m. Local dredging near the shoreline was done to locate the skimmer wall.
The discharge includes two offshore buried pipes conveying the flow from the plant to a point approximately 1,100 m at a seabed with elevation of –3.25 m. The two pipes are connected to a sheet pile box with a rectangular opening 7.5 m wide by 1.5 m high. The invert elevation of the opening is at –2.25 m (1 meter above the sea bottom), as shown on Figure 2. The top of the box extends above the maximum high tide and anticipated wave runup. The rectangular opening is designed to provide a surface discharge with uniform velocity at all tide conditions. This concept creates rapid initial mixing, thus reducing the size of the high temperature rise isotherms, as well as minimizing local seabed erosion.
Fig. 2 Outfall discharge box
The analytical studies encompassed all of the hydraulic and thermal analyses to size and select an alternative for use in developing the layout and in performing the thermal hydraulic physical model studies. The intake channel is designed with a flow velocity at low tide of 0.15 m/s to prevent scouring of the unlined channel bed and side slopes.
The thermal plume was modeled using CORMIX to determine its extent and to estimate the degree of warm water recirculation into the intake. The analysis made during ebb tide condition (flow toward the intake) showed an estimated warm water recirculation into the intake channel of less than 0.5 ºC. Figure 3 shows surface temperature rise isotherms for two-unit operation during ebb tide. The deep intake below the skimmer wall is designed with a low velocity of 0.15 m/s to minimize the withdrawal of floating trash, oil, and warm water from the surface layer. The exact depth of the opening below the skimmer wall was one of the objectives of the physical model study.
Fig. 3 Near-field thermal plume at low tide
Another objective of the analytical study was to determine the extent of erosion near the discharge structure. Two approaches were used. The first was to consider the heat decay from the surface discharge (as determined by CORMIX) as a measure for velocity decay. The second was to treat the discharge as a submerged discharge to maximize the velocity impact on the bottom. Using the finding from Albertson et al. [1] for submerged jets, the extent of the approximately 0.3 m/s velocity was used to initially select the extent of the riprap. The exact extent of erosion and the riprap to be used will be determined from the near field thermal hydraulic physical model.
Near-Field Model
The main purpose of the near-field model study was to investigate the detailed design of the discharge structure and to determine its optimal dimensions and alignment. The model represented a length of 840 m parallel to the shoreline of the gulf and 650 m normal to it. The discharge box and its buried pipes were also included.
The model study concentrated on the near-field of the cooling water discharge, where the inertia and buoyancy forces are predominant. Therefore, the model was operated according to the Densimetric Froude similitude relationship [3 & 5], and an undistorted geometrical scale of 1:25 was selected.
The model was designed to have a movable bed [2 & 4] to study the scour formation near the discharge structure, as well as to design the riprap protection. The model bed material selected was ionic resin, commercially known as “Amberlite IRA 955.” It has an average diameter, D50, of 0.6 mm and a specific weight of 1.08 t/m3.
The general test approach was to assess the adequacy of the discharge structure orientation and configuration and the proposed riprap size and extension. Various combinations of tidal currents and water levels for one- and two-unit operation were considered. For every run, surface horizontal plume temperature, vertical temperature distribution at three profiles, and velocity distribution near the discharge structure were measured. The scour hole that formed and the riprap movement were observed.
The temperature rise, DTx, at each point x was calculated using the following equation:
Where: To = outfall temperature, Ti = intake temperature, Tx = temperature at point x, and Tb= background temperature. Various combinations of tide water level, season, current direction and magnitude, and number of units in operation were simulated and tested.
The findings from this model study required the extension of the riprap apron to 187 m as compared with the 120 m estimated from the analytical studies. The selected riprap has D50 of 0.32 m. The configuration of the riprap apron is presented in Figure 4.
Fig. 4 Final design of outfall riprap apron
Far-Field Model
The main purpose of the far-field model was to prevent hot water recirculation. The studied aspects included distance between the intake and the outfall, alignment of the intake canal, and dimensions and location of the skimmer wall. The model represented a length of 6,200 m parallel to the shoreline and 1,800 m normal to it.
Due to the large area of the far-field, surface heat transfer is a major factor in heat dissipation and the resulting thermal plume and was considered in the model. Heat transport through the water surface is a complex mechanism in which radiation, evaporation, and convective transport are the major components. The excess heat loss from a thermal plume when the water surface has an artificially increased temperature can be conveniently described by linearizing the relation between the excess heat loss per unit surface area, Wh, and the excess surface temperature in the plume:
Wh = K(T – TE)
Where: Wh = heat loss per unit surface area (w/m2), K = heat transfer coefficient (w/m2 ºC), T = point temperature (ºC), and TE = equilibrium temperature (ºC). The value of the heat exchange coefficient in the prototype depends strongly on the meteorological conditions (wind speed, temperature, relative humidity, cloud cover) and can change considerably, even during the day. The heat exchange coefficient for laboratory conditions is usually lower than the prototype value due to the absence of wind in the model hall. The scale ratio for this coefficient can be related to the geometrical model scales. Analyzing the heat budget equation can derive this relation:
Where: Cw = specific heat of water, Qc = cooling water flow, and A = surface area. The left-hand term of this equation represents the total heat inflow into the water body, and the right-hand term represents the excess heat loss at the surface over the total area with elevated temperatures. With the scale ratios for the density (nQ) and the specific heat (ncw) equal to 1, one can get:
Now, it is clear that the far-field model must have scale distortion to reproduce the heat exchange through the water surface correctly. The selected scale was 1:100 horizontally and 1:20 vertically. Several cases were tested, including combinations of tidewater level, current direction, and number of units in operation. Typical surface temperature rise isotherms with current from NW to SE are shown on Figure 5.
Fig. 5 Far-field thermal hydraulic physical model thermal plume with current from NW to SE
Prediction of Prototype Temperature
To convert the measured temperatures in the model to those in the prototype, the following method was used. The one-dimensional conservation-of-energy equation gives:
Where: To = temperature at L = 0, TL = temperature at L, B = model width, L = distance between outfall and temperature measurement cross-section, K = heat loss coefficient, and Q = discharge. Applying the above equation to model and prototype obtains the ratio:
Determining DTr using model and prototype data, the model temperature data need to be adjusted by DTr to obtain prototype TL at distance L as follows:
The heat exchange coefficient can be calculated from various formulas on surface heat transfer such as the Meyer evaporation formula [6] as follows:
K = 4.5 + (b + 0.47)f (w)
Where: K = heat loss coefficient in watts/m2/ºC, b = vapor pressure coefficient, and
f(w) = 13.6 + 3.1U
Where: U = wind speed in m/s.
Warm Water Recirculation into the Intake Channel
The model study showed minor warm water recirculation with floating debris accumulation inside the intake channel. This finding required relocating the skimmer wall to the inlet of the channel but maintaining the total open area so that the velocity remained at 0.15 m/s. Warm water recirculation into the channel was determined to be negligible. This relocation necessitated modifying local dredging for the channel. The modified skimmer wall location and details are shown on Figure 6.
Fig. 6 Final configuration of skimmer wall
(1) The model showed that the extent of erosion near the outfall box is longer than analytically estimated.
(2) The initially selected orientation of the intake channel and outfall box is acceptable, and no measurable recirculation was observed.
(3) The possibilities of trash accumulation and oil inside the channel necessitated relocating the skimmer wall to the inlet of the channel.
(4) A distorted scale is required for the far-field model. Model distortion could be calculated from the prototype, the model heat transfer coefficients, and the length scale.
(5) The surface temperature rise isotherms in the model required correction to obtain prototype temperature rises.