Ghana - Thermal Power Project : environmental assessment (Vol. 4 of 7) : EIA : Takoradi Thermal Plant : addendum 2 revisions to EIA (final draft)

Environmental Assessment/Analysis Reports _ __ _ Report E0051 Ghana Thermal..power. Pro ect.. EA Category A- Environmental- Assessment August 1993: This r ra e p a b h o w o Co n - - : - : . - - . . .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ... 1 - -- - . ' :, , .:-, - . -,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ---- Thireprt as beenprepaed -:- by.. th Borroer o itsConsltan  Republic of Ghana Vofta River Authority Takoradi Thermal Plant Addendum 2 Revisions to Environmental Impact Assessment Final Draft January 1994 P10408.02 Acres Intemational Umited Niagara Falls, Ontario  Table of Contents Addendum 2 .....................................A2 -1 A2.1 Nobe MonHioing ..................................... A2 -1 A2.1.1 Methodology .................................... A2 -2 A2.12 Noise Results ................................... A2 -2 A2.1.3 Condusions and Recommendations ....... ............ A2 -7 A2.2 Air Quality Program .................................... A2 -7 A2.2.1 Air Quality Modelling ............... ............... A2 -7 A2.22 Air Quality Monitoring Program ........ .............. A2 -11 A2.3 Public Involvement Program ............. ................ A2 -21 A2.3.1 Community Involvement ........... ................ A2 -21 A2.32 Newspaper Advertisements ........ ................ A2 -22 A23.3 Distribution and Review of ELA ........ .............. A2 -22 A23.4 Public Meetings .................................. A2 -26 A2.3.5 Mechanisms for Land Acquisition and Payment of Cornpensation ........ .............. A2 -27 A.23.6 Land-Users at Proposed Thermal Plant Site ................................ A2 -2B A2A4 CherItem ......................... A2 -29 A2.4.1 Potable Water Supply ............................. A2 -29 Lst of Figures Lst of Tables  Lst of Tables Tab No. Tlu. A2.1 Results of Noise Measurements at Aboadze - Location NI A2.2 Results of Noise Measurements at Aboadze - Location N2 A2.3 Results of Noise Measurements at Aboadze - Location N3 A2.4 Results of Sulphur Dioxide (SO2) Sampling at Aboadze - Location Al A2.5 Results of Sulphur Dioxide (SO2) Sampling at Aboadze - Location A2 A2.6 Results of Sulphur Dioxide (SO2) Sampling at Aboadze - Location A3 A2.7 Results of (NOx) Sampling at Aboadze. Using Diffusion Tube Samplers L Location Al A2.8 Results of (NOx) Sampling at Aboadze, Using Diffusion Tube Samplers - Location A2 A2.9 Results of (NOx) Sampling at Aboadze, Using Diffusion Tube Samplers - Location A3 A2.10 Distribution of Environment Impact Assessment Report A2.11 Farmers on Acquisition Area at Aboadze  Environmental Impact Assessment Report Addendum 2 This Addendum to the Takoradi Thermal Power Plant Environmental Impact Assessment (EIA) provides an update of the investigations and actions which have been undertaken as input to the planning process for the plant since the main report was issued in August 1993. Addendum 1 was issued in September 1993 and covered revisions to specific items of text t is anticipated that the main report will be revised and issued in final form in early fall 1994. upon completion of specefic site investigatons. This Addendum resulls. in part, from requests by some lending agencies for more information on certain topics and presents the following information. - results of preliminary noise and air quality monitoring programs - public and agency input to the project - refined air quality modeling - update on refinements to plant and site features and potential impacts. A2.1 Noise Monitoring There was no previous information on background noise levels available for the EIA Concerns were raised that the proposed plant could provide elevated noise levels at nearby sensitive receptors, in particular, the local Islamic Primary School, located to the east of the proposed site. It was requested that ambient noise lvels be established, so that the existing noise environment could be profiled. A preliminary noise monitoring program was developed which the Environmental Protection Council implemented. This section outlines the program and provides the results.  A2 -2 A2.1.1 Methodology Three locations for noise monitoring were selected (Figure A2.1). These Included site Ni (near the Islamic Primary School); N2 (near the beach at Amuzu); and N3 (near the Methodist Church cemetery). These sites were selected as representative of the closest sensitve receptors, and the background evel at the site boundary. A Cirrus CRL 235A convertible Grade Sound Level Meter System was used. A- weighted readings were taken at Slow Setting. The sound level meter was fitted with microphone number MVI81A At each location measurements were taken within the following time slots: 0600 to 0800 hours 1400 to 1600 hours 1800 to 2000 hours 2200 to 0200 hours The time periods were chosen to represent working, relaxaton and sleeping conditions. Within each time slot, a continuous series of single measurements were taken at 5-second intervals until 100 readings were obtained. The surrounding conditions and unusual noise events were noted during the measurement period. Equivalent sound levels (Leq) were calculated from the series of dB(A) readings using the follovAng formula: Loq = 10^ L10 1) ]i 1N 0 + 10 + 101) where dBA1, dBA2 etc are the A-weighted sound level values oblained, and N = 100. A2.1.2 Noise Results The results of the noise measurements are provided in Tables A2.1 to A2.3 for readings taken between August 11 and 25, 1993. The results indicate that noise levels at all sites fall within the range of 42.3 to 56.1 dBA (Figure A22).  A2 -3 Site Ni lalamic Primary School The noise levels at the school were generally the lowest of all sites Investigated. Levels ranged between 42.6 and 49.5 dBA, with an average of 45B dBA. There was no consistency with respect to time of day for higher or lower nolse levels. The major sound sources were from birds, although the com mill was reported to be operational during certain time slots. It did not, however, appear to affect the noise levels measured. Sie N2 Near the Boech at the Community of Amuzu The average noise level of 49.4 dBA recorded at this site was the highest of all three sites, and levels ranged between 42.3 and 54.9 dBA fcr indMdual measurement sets. Interestingly the highest levels were recorded In the evening and during the nightfime. Although there are no remarks (see Table A2.2) provided for most of these readings of higher sound levels, they are probably related to the wind-generated noise experienced at this site such as the sound of waves breaking on the beach, and the rustling of the coconut palms. Some of the higher measurements also reflected rain events. The fishermen are typically active in the evening and boat engine noise was reported. The early morning (0600 hours) was the quietest time of day, when the highest frequency of calm conditions are experienced. Ste N3 Near the Methodist Cemetery The noise levels at this site showed the largest variation ranging from 43.3 to 56.1 dBA with an average value of 48.7 dBA. The quietest time of day was again early moming. This site is adjacent to the lnchaban-Aboadze Road and the higher noise levels were recorded midday and early evening, and are probably related to vehicle movement along the road.  A2-4 Table A2.1 Results of Noise Measurements at Aboadze Location NI (Near the School) No. Date and Timne Time Skot Leq (dBA) Remarks N=100 1 2Ig93 0600-0800 hours 45.6 Com mill on 0655-0705 hours Bird sounds 2 12/93 0600-0800 hours 46.0 Corn mill off 0706-0715 hours til 0712 3 12/8/93 1400-1600 hours 44.1 Bird sounds 1426-1435 hours Talding 4 12/8W93 1400-1600 hours 44.3 il ____ 1535-1544 hours 5 12/8/93 1800-2000 hours 49.5 1935-1944 hours 6 12/8/93 1800-2000 hours 46.7 1950-1959 hours 7 12/893 22.0040200 hours 42.6 221S-2224 hours 8 12f8/93 2200-20D hours 42.3 2230-2239 hours 9 21/8/93 2200-0200 hours 48.6 2330-2340 hours _ 10 2418/93 1400-1600 hours 48.1 Bird sounds 15.00-15.20 hours 11 25/8/93 0600-OB00 hours 46.5 Bird sounds 0730-0740 hours __--  Table A2.2 Resufts of Noise Measurements at Aboadze Locaton N2 (Near the Beach) No. Date and Time Time Slat Leq (dBA) Romwks N=100 1 11/8J93 1400-1600 hours 47.9 Bird sounds 1420-1440 hours 2 11/8/93 1400160M hours 48.4 Bird sounds 1450-1510 hours 3 12/8/93 0600-0800 hours 42.7 Bird sounds 0720-0729 hours ocean waves 4 12/a(93 0600-0800 hours 42.3 Bird sounds 0730-0745 hours ocean waves 5 -2. 1800-2DOO hours 50.7 Boat engines 1825-1835- hours Wind noise 6 12/8/93 1800-2000 hours 51.5 1837-1850 hours ! - 7 12/8/93 2200-C0 hours 54.9 ;_ ZOO2200-223D hours 8 1218193 2200-0200 hours 542 2240-2310 hours _ _ 9 16/8(93 0600-0800 hours 47.7 j 0740-0750 hours _ 10 16/8J93 0600-0800 hours 46.5 0750-0OOD hours 11 16/W93 1400-1600 hours 51.9 Wind and Rain 1400-1410 hours . I 12 16/B/93 1400-1600 hours 51.5 Wind and Rain 1420-1430 hours 13 17/8193 1800 -2000 hours 50.9 1830-1845 hours l 14 17/(93 1600-2000 hours 50.5 1900-1915 15 21/8193 220002 hours 52.6 2230-2240 hours l 16 23/8193 1400-1600 hours 48.5 1400-1410 hours l 17 259 0600-080D hours 47.4 0750-0800 hours  A2 -6 Table A2.3 Results of Noise Measurements at Aboadze Locadon N3 (Near the Cemetery) No. Date aN Tnme Tim. Slot Leq (dBA) 1 Remark N=100 1 13/93 1400-1600 hours 542 Vehicular noise 1400-1420 hours 2 13(193 1400-1600 hours 56.1 Bird sounds 1430-1500 hours Vehicular noise 3 13893 180-D2000 hours 50.4 Bird sounds 1900-1920 hours 4 13(8/93 1800-2000 hours 50.8 Bird sounds 1930-1950 hours 5 14/893 060800 hours 46.7 Vehicular noise _______ 0730-0740 hours Bird sounds 6 141193 0600400 hours 43.3 0750-0800 hours 7 17/8/93 1400-1600 hours 45.3 Vehicular noise 1545.1600 hours . Bird sounds 8 21/1893 22.00-0200 hours 47.9 Crickets from 2400-2410 hours N=1 to N=25 9 258/93 oBo6-o0 hours 44.0 Vehicular noise 0710-0720 hours Bird sounds  A2 -7 A2.1.3 Conciusions and Recommendations In Section 5.2.2.1 of the EUA (Noise and Vibration) the predicted noise level due to the proposed plant operation at the closest sensitive receptor (i.e. the primary school) was 49 dBA. This level Is marginally higher than the existing conditions as measured (the arithmetic mean of the sound level measurements was 45.8 dBA). The predicted noise level In the E-A was obtained without factoring In any attenuation produced by the use of noise barriers between the site and the school. Screening of the site has been recommended for aesthetic, as well as noise reduction purposes. These noise levels should be acceptable and not cause any disruption to teaching actMtles Inside the school. A further noise monitoring program has been recommended. This new program is to be a repetition of the preliminary investigation with a few modifications. It will also include more observations regarding weather conditions and nolse-making activties; and to include one more monitoring site. The location.of the permanent townsite has now been finalized (a site immediately north of the proposed thermal plant site, north of the road) and it is recomnended that background noise levels at this location be determined. One time period in the daily program has been changed; the 22.00 to 02.00 has been reduced to 00.00 to 02.00 to more closely reflect sleeping conditions. A2.2 Air Quality Program A2.2.1 Air Quality Modelling Modelling undertaken and presented in the EIA was conducted with the Ontario Regulation 308 Air Dispersion Model. This model, while effective at predicting worst case Ground Level Concentrations (GLC). under a variety of preset meteorological parameters does not take into account local meteorological conditions. Two model runs were however presented in the EIA [Appendix E. Figures E21 (a2) and a(3)]. in which the wind speed was fixed at a low velocitv (2.235 mlsec) considered to be somewhat representative of actual Ghanaian flild conditions. These runs indicated that the point of maximum plume impingement would be located 15.20 km downwind of the plant, under those circumstances. Thus, a series of model runs have been subsequently undertaken with the US Environmental Protection Agency (EPA) Industrial Source Complex (ISC) model to more accurately define the location of the point of maximum impingement, and to refine predictions of GLCs associated with this maximum. This modelling has also  A2 -B been used to select appropriate monitoring sites for the long-term air quality monitoring programme. Meteorological data from Takoradi Air Force Base (Annex 1) has been used as input to the model. For runs with the iSCST2 (Short Term) Model, the range of potential atmospheric stabilities was revised according to wind speed as presented below. Wind speed Possble StabilEy (mis) Cum c 3 A, B, C, D, E. F 3-4 B,C,D.E 4-6 C, D >6 C,D. The range o1 possible atmospheric stabilities changes with wAnd speed; at higher wind speeds the exdreme stabilities cannot occur because of the strong mixing. Due to the extremely light wind speeds in the project area, GLCs in the immediate vicinity of the plant were overestimated by the previous Ontario Regulation 3OB model (i.e. undertakes a series of iterations at presei wind speeds (1 - 18 m/sec) and stability classes), and the previous exercise was undertaken to improve the predictive capability of the ISC short-term model.for the Ghanaian situation. ISC2 Long-Term ModelUng Modelling runs were undertaken to assess long-term GLCs associated with a number of operational scenarios and plant configurations, which include separate single stacks and a single multi-flue stack. Runs were also undertaken to assess the impact on local air quality of operating in bypass mode prior to water being available for NO, control. The emission rate for two unit controlled (i.e. NO, control in operation) operation was set at 102 gmisec. which represents 51 gm/sec/unit. This value was obtained following a review of emission characteristics of representative Combustion Turbine Generator (CTG) units, and included a factor for fuel-bound nitrogen (FBN), based on characteristics of a West African lighl crude oil. A fuel of this nature will be used to fire the plant. Modelling was based on a stack height of 40 m, a stack diameter of 6 m, and exhaust gas exit velocitie_ "f 332 and 18.1 rrVsec, for bypass and combined cycle operations, respectively. This represents exhaust gas temperatures of 554 and 175C for bypass and combined cycle operations, respecfively. The various operational scenarios which were modelled with their associated stack configurations are presented below.  A2 -9 Emibson Un Muli- Rats Unit in NO. Units GI Cornbied Sepaate flue Figure Conilguration (g/s) Operation Control Byas Cycle SackBs dack No. A 102 2 yes 2 X A2.3 B 102 2 Ye 2 - X A2.4 C 102 2 yes 2 X A2.5 D 102 2 Yes - 2 X A2.6 E 157 1 No 1 - X A2.7 | F | 314 2 No2 - I - I X A2.B l The results of the modelling exercise are presented graphically as contours of GLC of NO, (expressed as ig/nm3) overlain on a project area map in Figures A2.3 to A2.8. while computer output from individual model runs is presented in Annex 1. As noted above, each figure represents a specific set of potential operational conditions. The initial observation that can be made from examination of Figures A2.3 to A2.6. is that the multi-l-ue stack configuration provides substantially lower GLCs within the realm of predictive capability of the model (50 km in each direction from source) than the separate stack configuration under both bypass operations (Figure A2.3 vs Figure A2.4) and combined cycle operations (Figure A2.5 vs Figure A2.6). This is primarily due to the greater mass of the plume originating from the multi-flue stack, which exits as one large mass of hot, exhaust gases compared to two smaller masses. The larger plume having more momentum and buoyancy rises higher and travels further betore retuming to ground level. The figures also illustrate the point of maximum impingement and ground level concentrations associated wIth this maximum for the various scenarios. If a 40 m multi-flue stack Is utilized for the Takoradi Thermal Plant, the point of maximum impingement under normal base load operations (i.e. combined cycle) would be 10 - 14 km NNE of the plant behteen the villages of Daboasi and Beposo (Figure A2.6). The long-term annual average increment to GLC of NO, is predicted to be 25 - 31 pgM3 in this area. Concentrations decline quicidy to the south and west, resulting in an average annual Increment to GLC of NO. at Aboadze of 5 -10 pg/M3. Given the low background concentrations measured in the project area (< 7 ig/r3, see Section A222). the resulant long-term average concentration is well below estabrished limits (100 pg/M3 annual average, World Bank) and in line with those spec'ied by ADB's appraisal mission representatives (30- 35ig/rn3 annual average).  A2 -10 GLCs predicted for bypass operation are considerably iower (see Figures A2.3 and A2.4) than for combined cycle operation, and would add less than 11 pg/M3 to ambient NO, concentrations under the separate stack scenario, which Is the envisioned configuration for the plant. A multi-1lue stack scenario (Figure A2.6) is also presented for comparison. In order to evaluate the potential Impact of plant operations prior to the availability of water for NOx control, model runs were undertaken representing one and two unit, uncontrolled bypass operation. One unit operation results In a maximum Increment to GLC of NO, of 15 - 16.2 pgIM3, centred around the village of Beposo, while two unit operation can be estimated by multiplying the contour values presented In Figure A2.3 by 3.08. Thus, under two-unit, uncontrolled, bypass operatlon, with 40 m stacks, the annual average Increment to GLC of NOx at the point of mamdmum impingement would be approximately 31 pg/m3, still well within the recommended values. This would be located in the vicinity of Beposo, on the Takoradi-Accra highway, just east of the Pra River. Detail design approaches may further Improve this conditlon. ISC2 Short-Term Modelling The ISCST2 Model was utilized to estimate 1-h maximum Ground Level Concentrations of NOx associated with the six configuraiions (A - F) presented in the table in the preceding section. The results of these model runs are presented graphically in Figure A2.9 and presented in tabular form in Annex 1. These runs indicate that, under the anticipated base load operational conditions (configuration D), 1-h maximum GLC would be less than 31 1igIm3 under all stablity classes, and would occur between 1 and 2 km from the plant. This GLC compares favourably with that obtained with the long-term modelling. However, one must temper results with the knowledge that they represent worst case conditions, which were generated by preset model wind speeds, which are not entirely representative (i.e. higher) of the project area. This modelling does however serve to emphasis that the values produced by the ISC2LT Model are reasonable, and that the multi-flue stack conliguration produces lower GLCs than can be achieved with single stacks. A second parameter examined under ISCS12 Model runs is the effect of stack height on 1-h maximum GLCs. Configurafion D parameters (i.e. two unit, controlled, combined cycle operation, multi-flue stack) were maintained throughout a series of model runs, while stack height was varied from 20 to 100 m, in 20 increments. The resultant plot of Ground Level Concentrations of NOx for the various stack heights vs distance from the stack is presented diagrammatically in Figure A210, and in  A2 -11 tabular form In Annex 1. As noted, a stack height of 40 m produces a maximum Increment to GLC of NOx of approximately 31 pg/M3 at a distance of 2 Ikm from the plant. Again It should be noted that the distance of maximum GLC from the plant is not considered accurate, due to the low project area wind speeds vs the model wind speed which produces the worst case scenario. The figure does however Indicate the effect that stack height has on ground level NO, concentrations. Sulphur Dioxide (S02) Although not specifically modelled, GLC of S02 can be Interpreted from the existing NOx figures and tables by applying an appropriate conversion factor to obtain a corresponding GLC of S02. Assuming that the plant emission rate, with luel of 0.2% sulphur, is 67 gm/sec for two unit operation, the factor to convert NO. GLC to SO2 GLC would be 0.66. Thus, for the base load, combined cycle operational conditions as represented by configuration D, the GLC of SO2 in the area of maximum impingement would be 20.5 pg/m3. Concentrations in the vicinity of Aboadze would range from 3 - 6 pg/m3. The shape of the contour plot would be similar to that presented in Figure A2.5, however values shown on each contour line would require multiplication by 0.66 to represent GLC of S0. As with the NO, plot, the area of maximum impingement is located between Daboasi and Beposo. Under 2-unit bypass operations, the area of maximum GLC is centered about Beposo (see Figure A2.3) wAth values ranging from 6 - 7 pg/m3, and declining to 1 - 2p/m3 in the vicinity of Aboadze. Thus, under both operational conditions (combined cycle and bypass), GLC ot SO2 are well below the incremental loading rate guideline of 50 pg/mr3 for unpolluted conditions. which is the case for the project area (see Section A222). A2.2.2 Air Quality Monitoring Program When preparing the EIA there was very limited information available on air quality in the Aboadze area and in particular there was no data relating to the components which could be impacted by the plant development, namely SO2 and NOx. It is necessary to obtain an understanding of the ambient conditions in order to be able to assess if there have been changes in ground level concentrations of pollutants in the future, resulting from plant operations.  A2 -12 A preliminary air quality program was undertaken In order to meet these deficiencies and to fuMlill requirements of lhe lending agencies. This section describes the program and the results obtained. The program, developed by Acres. was also executed by the Environrnental Protection Council (EPC). It was adapted somewhat by EPC to more closely meet the conditions at the site and the equipment available. The following summarizes the Inlormation provied in the EPC October 1993 preliminary 'Report of Air Quality Sampling at Aboadze near Sekondi for Background Noise Levels, Concentratons of Sulfur Dioxide and Nitrogen Oxides in the Ambient Air'. MtdbOxdogy Air quality was monitored at three locations (see Figure A2.1). - A control location, west of the proposed site, on the beach ridge approximately 100 m inland and to the west of the small settlement (5 to 8 houses) in this vicinity (ocation Al). - On the access road, at the northeast corner of the site, near the cemetery (xocation A2). - Approximately mid-way along the Aboadze-Aboese road, on the north side of the road approximately 100 m north of the road (location A3). At each location SO2 and NO, were measured. Ozone (03) was to be included in the program, but EPC was not equipped to undertake its monitoring. The proposed monitoring program time frames were as follows. - As a one time event to determine daily variation, one reading per hour, during each hour, over a 24-h period. - On a weekly basis, one day per week. 4 times per day, at 0800.1200,1600 and 2000 hours. S02 was measured using a Casella sulphur dioxide sampler. Hourly and half hourly measurements were taken. SO2 can be measured directly in the .<:ld with the sampler, the concentration being directly related to a change in conductivity of an electrolyte.  A2 -13 Gradco Diffusion Tubes (DIF 100 RTU) were used to measure the concentration of NO, (NO+NO2) in the ambient air. The tubes were mounted on a post above ground level. The NOx sampling program was redesigned by the EPC after the first laboratory analyses when they were unable to detect any NO, after 7 days of exposure. Sampling periods for NO, were subsequentiy lengthened to 30 days, or more, in order to obtain detection. Detailed Informafion on the S02 and NOx sampling equipment, its calibration and the formula for determining concentrations is provided in Annex 2 (EPC. October 1993). Results The results of the SO2 and NOx sampling are provided in Tables A2.4 - A2.9. SO2 S02 monitoring was predominantly undertaken for 1-h sampling times using the 0800. 1200 and 1600 time slots only. The S02 results for each site are summarized below. Locaion Range Averugo (g/r3) Concentraon^ (ug/m3) Al (near the beach) Not detected - 10 23 A2 (near the cemetery) Not detected - 160 35 A3 (near Aboesi) Not detected - 7 4 Although it is very difficult to see any pattems in so few results, the lowest SO2 levels were recorded in the early moming and near Aboesi. The higher results may reflect local fires/fish smoking operations. Amea cocetaion as detemw t d by talkng a alue odl pgW fbr the 'non-etected' events.  A2 -14 NO3 As noted above, only after sampling over 700 hours was It possible to detect NOx in the atmosphere with the equipment available. The NOx results for each site are summarized below. Localtn Rang (pg/rn3) Al (near the beach) 3.4 - 5.0 A2 (near the cemetery) 1.9 - 5.1 A3 (near Aboesi) 1.9 - 7.0 Conl and R