Impacts of geothermal well testing on exposed vegetation in the Northern Negros Geothermal Project, Philippines

Josefo B. Tuyor, Agnes C. de Jesus*, Reinero S. Medrano, Jo Rowena D. Garcia,

Sherwin M. Salinio and Leonora S. Santos

PNOC-Energy Development Corporation, Merritt Road, Fort Bonifacio, Taguig, Philippines

Received 3 February 2004; accepted 30 September 2004 Abstract The impacts of geothermal fluid discharges during the testing of Pataan 5-D well were evaluated on seedlings of mahogany (Swietenia macrophylla King) at various distances from the well and on natural forest vegetation around the wellpad. Parameters measured were: 1) geothermal brine spray concentration, 2) plant concentration of geothermal signature ions (B, Cl, Li, and Na), 3) symptoms of plant damage and 4) plant recovery. Meteorological parameters were also gathered. Adverse effects on the test plants were observed at distances of 5-50 m from the well silencer for over-spray during the horizontal discharge and at 50-350 m from the wellhead during the vertical discharge. Salinity was identified as a significant cause of plant damage. Observed symptoms of damage included drying of leaf tissues expressed as necrotic areas, which occurred first at the tip of older leaves and progressed along the margins as severity increased resulting in abnormal defoliation. Recovery of seedlings and natural vegetation from sprays was observed in both vertical and horizontal discharges. Keywords: geothermal discharges, environment, vegetation, defoliation, salinity 1. Introduction Well testing is conducted during the exploration and development stages of a geothermal project to characterize the physical and chemical properties of the produced geothermal fluids, the permeability and fluid state of the reservoir, and the power potential of the wells. During well testing, geothermal fluids are often released into the atmosphere at relatively high temperatures and pressures. The fluids contain salts (mainly sodium chloride) and elements such as arsenic, boron and lithium, which may have detrimental effects on sensitive plants that have been in contact with geothermal spray. However, there is little published information on the effects of geothermal discharges on vegetation, despite its importance in forested geothermal areas where there is a growing concern for biodiversity and life-support systems in the environment. _______________________________________________________ * Corresponding author. Tel. +63-2-893-6001; fax: +63-2-817-9154 E-mail address: dejesuac@energy.com.ph

1

This study was conducted to: 1) document, characterize and evaluate the impacts of both vertical and horizontal discharges during testing on exposed vegetation, 2) identify the causes of damage, 3) determine the threshold of exposed plant species as regard to the levels of relevant chemical elements, and 4) determine the allowable duration of discharge without mitigation that will not cause death to exposed plants. 2. Methodology 2.1 Study Site The study was conducted in the vicinity of well PT-5D located in the Pataan sector of the Northern Negros Geothermal Project in Negros Occidental (Fig. 1). The well was completed in February 2001 and has a vertical depth (vd) of 2250 m. Produced fluids (steam + neutral-pH brine) are mainly from a depth of about 2150 m (vd) where the reservoir has a temperature of about 290 oC. When fully opened, the well discharges at a rate of approximately 60 kg/s. 2.2 Experimental Set-up The experiment was done during the testing of well PT-5D from May-June 2001. Locally grown, 5-8-month old mahogany (Swietenia macrophylla King) seedlings, and natural vegetation (mostly pioneer species) were used as test plants. The mahogany seedlings were distributed along the well pad radius (0-50 m) and in open areas beyond the pad (50-200 m), as permitted by the terrain for both horizontal and vertical discharges. In addition to the mahogany seedlings, a strip of natural vegetation starting at 50 m from the well was included in the physical damage observation and other analytical tests. Unexposed seedlings established at the nursery (about 1.5 km from the study site) and the natural vegetation at an unaffected pad (Pad A, about 1.2 km from the study site) served as the control. Catch basins were placed alongside the test seedlings in each station to collect the geothermal sprays for analysis of geothermal indicator ions such Cl, Na, B and Li. As a mitigating measure, in forest areas, an inclined exhaust spool was fabricated to ensure that the plume of the geothermal spray would point away from the forested area. The exhaust spool was deliberately adjusted to provide a NNW plume direction, where the forest was least dense. For vertical discharge experiments, five lines were established radiating from the wellhead a) along the discharge plume direction (NNW), b) opposite the discharge plume (SSE) and c) quadrants left and right of the discharge plume (NW, NE, E). In each line, stations were established at distances of 20, 50, 75, 100, 125, 150 and 200 m from the wellhead. Four (4) replicate seedlings of mahogany were exposed in each station. Two (2) seedlings were used for observation of physical damage and tissue analysis of geothermal indicator ions, and the other two (2) replicates were observed for signs of recovery. For the horizontal discharge experiments, eight lines were established because greater dispersion of the spray was expected; these lines were oriented at the compass quadrants (i.e., N, NE, NW…). In each line, six (6) stations were established at distances of 5, 15, 30, 50, 75 and 100 m from the wellhead. Four (4) replicate seedlings of mahogany were exposed in each station, similar to the vertical discharge experiment. In addition, control seedlings and natural vegetation were also assessed.

2

2.3 Discharge period As a standard procedure, the company (PNOC-Energy Development Corporation) conducts vertical discharge tests for a shorter time period than for horizontal discharge testing because of the greater dispersion capacity of the geothermal overspray because of the greater fluid exit pressure when the wells are discharged vertically. The horizontal discharge tests were conducted near-continuously from May 16 to June 30, 2001 for a total of 45 days. The vertical discharge tests were conducted on five separate occasions at weekly intervals (May 29, June 6, June 14, June 22, and June 29) for a total of 10 hours, or about 2.5 hours/discharge. During this time, the discharge rate varied between 35 and 60 kg/s.

2.4 Schedule of sample observation, collection and harvesting For the vertical discharge tests, sample collections started only 90 minutes after each discharge due to zero visibility at the pad area during the first 60 minutes of discharge. For the horizontal discharge tests, collections of seedlings, leaves of natural vegetation and catch basins at distances 5, 15, and 30 m were done every three days during the first 14 days of discharge. Collection frequency was later reduced to every five days, then every 15 days (June 30), for a total of nine collections. Samples at 50, 75 and 100 m were collected less frequently during the first 30 days of discharge (i.e., May 21, 31, June 10, 15), which was later adjusted to every three (3) days as discharge went on (i.e., June 18, 21, 24, 27, 30). The difference in the collection schedule of various stations/distances was based on the premise that vegetation close to the source of discharge was more exposed and thus may manifest damage at a much faster rate. The opposite is true for vegetation far from the source. 2.5 Meteorological conditions Meteorological parameters, which may have a significant influence on the degree of impact, were recorded. These consisted of ambient air temperature, wind direction, wind speed and rainfall. Except for ambient temperature, which was measured within a 5 m radius of the silencer and at the forest edge (about 50 m from the silencer), the rest of the parameters were measured in randomly selected stations. Readings were taken daily at 8 a.m., 11 a.m., 2 p.m., 5 p.m., except for rainfall, which was read only once a day at 5 p.m. These parameters are relevant in assessing the final chemical levels of geothermal sprays in the catch basins, test seedlings and natural vegetation. 2.6 Observations of damage on test seedlings and natural vegetation The manifestations of damage on the test seedlings, as well as on the natural vegetation, were observed prior to leaf harvesting for chemical analysis. For the natural vegetation about 10 leaf samples were observed for physical damage, while for the test seedlings all leaves were observed. The symptoms and area of leaf damage per seedling per station were noted qualitatively and photographs were taken. 2.7 Chemical analysis of test seedlings and natural vegetation

3

Leaves of both test seedlings and natural vegetation at each station were washed off with 100-ml of de-ionized water and the washings were analyzed for geothermal indicator ions such as B, Cl, Li and Na. After washing off, the leaves were collected and analyzed for the same indicator ions.

2.8 Soil contamination

The contribution of systemic absorption, through the roots, was assessed by measuring soil contamination. The soils from the exposed pots were analyzed for B, Cl, Li and Na.

2.9 Recovery of defoliated test seedlings and natural vegetation As defoliation may mean only dormancy, the live status of the test seedlings and natural vegetation was tested by exposing the cambium of the bark to 3% tetrazolium chloride solution. About 10 spots were tested in each test seedling and natural vegetation. Positive test up to 50% indicates live status. Other signs of recovery such as foliage formation were also noted. This was done for three months. 2.10 Chemical analysis of geothermal fluids The profiles of the potential causes of damage, such as the daily concentration of geothermal ions at the weirbox and fluid temperature, were recorded. In addition, the geothermal brine sprays collected in the catch basins at each station were collected at the same time the seedlings were harvested. The brine was washed off from the basin with 300-ml of de-ionized water and the washings were analyzed for geothermal indicator ions. 3. Results The horizontal discharge of well PT-5D was conducted for 45 days and the vertical discharge was conducted for a total of 10 hours or approximately 2.5 hours/discharge, as indicated earlier. The impacts of vertical discharge in the present study represented the worst scenario because the standard vertical discharge for Philippine wells only lasts for 30 minutes at any single time.

3.1 Meteorological parameters The meteorological readings at the study site during the discharge are shown in Table 1. The ambient temperatures at 5 m from the silencer and at the forest edge (>50 m from the silencer) were between 23–25 oC, indicating temperatures were near uniform throughout the survey area. The dominant wind directions during the whole test period were from the northeast and southeast, but the northeasterly winds predominated during the entire time of the discharges. Other wind directions recorded were southwest (SW), south (S), east (E) and northwest (NW). Average winds speeds and total rainfall in May and June were similar, although the amounts of rainfall recorded during the discharge periods were greater than

4

normal for the area. Because of the inclination of the exhaust spool, the resulting plume direction, rather than the prevailing wind direction, affected the observed impacts on vegetation. 3.2 Chemical analysis Table 2 shows the concentrations of B, Cl, Li, Na and pH at the weirbox. The pH of the well fluids was generally neutral (av. pH = 6.96) because the well was acidized prior to its discharge. This neutrality therefore eliminates pH as a possible cause of damage. The concentrations of B, Cl, Li and Na in the geothermal sprays collected from the catch basins decreased with distance from the silencer (Table 3, Fig. 2). The highest concentrations were recorded within the 5-m radius. The same trend was also recorded from the washings (Table 4) and the leaves of the test seedlings (Table 5), and from the exposed soils (Table 6). In contrast, ion levels in the control remained low. Exposed natural vegetation also manifested elevated ion levels during the discharge compared with the low control and pre-discharge levels (Table 7). 3.3 Effects of horizontal discharge on test seedlings and natural vegetation Early damage symptoms assessed one day after the discharge included drying of leaf tissues or dessication observed as necrotic areas. This occurred first at the tip of older leaves and progressed along the edges or margins as severity increased, resulting in abnormal early defoliation on the secondday after discharge (Fig 3). The impact zone was confined to within a 50 m radius of the silencer, regardless of the wind direction. Test seedlings at 5-30 m from the silencer manifested symptoms one day after the discharge, while those at 50 m manifested five days later. Natural vegetation was not affected during the horizontal discharge. No impact was observed in seedlings exposed at distances of 75-100 m. In terms of extent of plant damage and recovery, seedlings within 5-15 m from the wellhead were severely affected compared to those seedlings at 30-50 m. . Test seedlings exposed to the horizontal discharges had varying degrees of recovery (Table 8). Severely affected seedlings at 5-15 m distance and those with longer exposure registered low recovery rates. The reverse was noted for those exposed at 30-50 m and those with short exposure. The overall recovery rate of seedlings exposed to horizontal discharge was about 70%. 3.4 Effects of vertical discharge on test seedlings and natural vegetation Due to the angle of dispersion, the impact zone of vertical discharges started only at 50 m (i.e. beyond the 50 m radius of the pad) and only in the plume direction (NNW) and its immediate left (NW) and right (NE) quadrants. Thus, only the natural vegetation within this zone from 50 to 350 m from the wellhead was affected. Damage symptoms manifested were similar to those observed intest seedlings affected by the horizontal discharge, except that defoliation started only on the sixth day after the first vertical discharge. In contrast to the horizontal discharge, the natural vegetation exposed to vertical discharge registered 100% recovery three months after discharge. This higher recovery index is due to the shorter exposure period to vertical discharges. Test seedlings exposed to the horizontal discharge had varying

5

degrees of recovery (Table 8). Severely affected seedlings, at 5-15 m distance and those with longer exposure, registered low recovery rates. The reverse was noted for those exposed at 30-50 m, and those with short exposure. The overall recovery rate of seedlings affected by horizontal discharge was about 70%. In contrast, the natural vegetation exposed to vertical discharge registered a 100% recovery three months after the discharge (Figures 4a – 4d). 4. Discussion There is a dearth of information on the effects of geothermal sprays on vegetation. The results of the study can serve as reference in developing measures to prevent or minimize the adverse impacts on vegetation. This is important in forested geothermal areas in the light of the growing concern for biodiversity and life-support systems in the environment. 4.1 Causes of Damage In the present study, three potential vectors of damage were assessed, namely: temperature, salinity and toxic ions. Temperature can be ruled out as a cause because ambient temperatures measured within and outside of the impact zones were not significantly different. Defoliation also occurred at 50 m and beyond from the wellhead, where ambient temperatures are similar to those recorded at the impact zone. The primary cause of damage, based on symptomatic manifestations, is salinity. Early symptoms of affected plants, including drying of leaf tissues, which occurred first at the tip of older leaves and progressed along the edges or margins as severity increased, are signature symptoms of salinity (Ayers and Westcot, 1976; Peacock, 1998; Seawell and Agbenowosi, 1998; and Futch and Tucker, 2001). Excessive leaf drying is often accompanied by abnormal early defoliation (Ayers and Westcot, 1976). According to Alsup (1998), salinity affects plants in three major ways: water deficit, ion toxicity and nutrient imbalance. This study seems to support the importance of the first effect. The dessication of leaves indicates water deficit due to the withdrawal of water following the water concentration gradient (high water concentration in the plant cell to the low water concentration in the saline geothermal spray). On the other hand, there were no signs of black patches that are typical of toxic ion dumping on cells. Salinity as cause, rather than toxicity, is consistent with the observation of the damage manifested in older leaves as against ion toxicity, whose prime targets are the younger leaf tips which have a greater water fraction. Deaths in younger shoot tips appearing as black spots, which are due to toxic ions, were not observed in the study. Table 9 shows the tolerance limits of plants to salinity, which have been exceeded during this the study. One of the toxic elements in the geothermal spray (i.e. Li) seems not to be a factor in plant damage because the levels detected are below those that may cause plant injury (Schauss, undated). However, the parameters measured were insufficient to assess mineral nutritional effects. 4.2 Possible mode of uptake Although the concentrations of ions in soils are high, the main cause of vegetation damage appears to be the direct contact of leaves with ions from geothermal sprays because of the

6

immediate dessication response of the test plants (observed as early as the first day after the discharge). The saline ions of Na, Cl and B may have been absorbed directly by the test plants through the cuticular openings of the leaves. This was confirmed by the presence of salt deposits on the leaf surfaces of test plants, especially at distances of 5-15 m. Uptake from the root systems may have taken place for test plants with prolonged exposure, and this may be one of the reasons why recovery rate was very low. 4.3 Threshold of the exposed vegetation The combined LC50 (i.e. the lethal dose that will kill >50% of the test seedlings) of the three saline ions measured, is at 8,000 ppm (B=79ppm; Cl=5200ppm; Na=2900ppm). This was recorded within the 5 m radius from the silencer. Other ion contributors need to be studied, including the kinetics of ion transport to determine what causes the decrease in ion levels with distance from the well. 4.4 Recovery of affected vegetation The recovery of affected plants was dependent on the concentration of geothermal ions absorbed from the sprays, and the duration of exposure. As expected, test plants close to the silencer, which were exposed to higher ion levels, showed less recovery than those farther away (<50%). The same was true for test plants with prolonged exposure to discharge, presumably due to their continuous exposure and possible ion uptake through the roots. Recovery was also dependent on age of the affected plants, with seedlings recovering poorly compared to the natural vegetation. According to Nelson (1991), seedlings are more sensitive to salts than established plants. This may be due to the capacity of mature plants to buffer stresses or correct altered processes. The drying up of plant tissues and cells is also expected to cause the alteration in ultra-structures or the cell organelles, which are the seat of important life processes. Irreversible alteration leads to death of the plant parts (Cortez, 1978). The present study confirms the earlier observations noted in the other of the company’s geothermal projects ,i.e Mt. Labo (Mindanao) and Mt. Apo (Leyte), that the impacts of geothermal discharges during well tests are temporary. In these fields, defoliated vegetation fully recovered in a range of one week to years after the completion of the well tests(Fernandez, 1993). 5. Conclusions The results of the study indicated that one of the possible causes of vegetation damage during geothermal well testing is salinity, associated with high concentrations of saline ions in the geothermal brine sprays released from the discharging well. It also showed that the impact zone of horizontal discharge is confined within a limited area around the well compared with that of the vertical discharge. The horizontal discharge affected only the pad area. Thus, vertical discharge was more harmful to vegetation than horizontal discharge since the wellpad area is normally cleared of vegetation. It was also determined that the severity of vegetation damage is dependent on 3 factors: 1) resulting brine spray concentrations, 2) distance from the

7

discharging well, and 3) prevailing wind (plume) direction during the discharge. The study also confirmed that the impact of the well testing on vegetation was only temporary. 6. References Alsup, C. M., 1998. Salinity: causes and effects, and management practices. Sourced from the

internet at http://www.geocities.com/clydealsup/Salinity.htm. Ayers, R. S., Westcot, D. W., 1976. Water quality for agriculture. Irrigation and drainage paper

29, Food and Agriculture Organization, Rome, Italy. Cortez, A. R. 1978. The growth responses of Phaseolus vulgaris L. cv. “White Baguio” to low

water status. Unpublished M. Sc. Thesis, Univ. of the Philippines, Diliman, Quezon City, Philippines.

Fernandez, C. S. 1993. Hydrogen sulfide and sulfur dioxide from geothermal sources: its effect

on forest vegetation. Unpublished M.Sc. Thesis, Univ. of Canterbury, Christchurch, New Zealand, 139 p.

Futch, S. H., Tucker, D. P. H., 2001. A guide to citrus nutritional deficiency and toxicity

identification. Horticultural Sciences Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, 32611, USA.

Mathers, H., undated. Cyberconference: water quality for woody plants. Oregon State

University, USA. Nelson, P. V., 1991. Greenhouse operation and management. 4th ed. Prentice Hall, Englewood

Cliffs, USA. Peacock, B., 1998. The use of soil and water analysis. University of California Cooperative

Extension, Tulane County, California, USA. Schauss, A. G., undated. Lithium. Sourced from the internet at

http://www.sonarm.com/research/lithium.htm. Seawell, C., Agbenowosi, N., 1998. Effects of road de-icing salts on groundwater systems.

Groundwater Pollution Primer CE 4594, Soil and Groundwater Pollution, Civil Engineering Department, Virginia Polytechnic Institute and State University, Virginia, USA.

8

Table 1 Meteorological readings at the study site; average of 4 readings per day per month.

Month Ambient Temperature (oC)

Wind Directiona

Wind Speed (m/s)

Rainfall (cm)

5 m radius >50 m May 2001 24.45 24.52 NE 1.27 2.94 June 2001 23.96 24.50 NE 1.24 3.14 a Predominant wind direction based on 4 readings per day per month. Table 2. Concentrations of B, Cl, Li, Na and pH in the weirbox; average of the entire 45-day discharge. pH Boron

(ppm) Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

6.96 259 15,542 33.92 8,215 Table 3 Concentrations of B, Cl, Li and Na in the geothermal spray collected by the catch basins during horizontal discharge; average of 9 collections. Distance from the silencer (m)

Boron (ppm)

Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

5 136 9085 11 4884 15 97 5972 7.5 3186 30 65 4561 9.5 1728 50 65 4112 6.2 1980 75 37 2498 5.7 1319 100 23 1548 2.2 820

9

Table 4 Concentrations of B, Cl, Li and Na in the washings of the test seedlings; average of two collections and 6 replicates/collections. Distance from the silencer (m)

Boron (ppm)

Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

Control <0.10 2.95 <0.01 0.45 5 0.25 26.67 0.05 11.71 15 <0.10 8.23 0.02 3.53 30 <0.10 5.06 0.01 1.38 50 <0.10 3.06 <0.01 1.23 75 <0.10 2.92 <0.01 0.93 100 <0.10 2.57 <0.01 0.53 Table 5. Concentrations of B, Cl, Li and Na in the leaves of test seedlings exposed during horizontal discharge; average of 9 collections. Distance from the silencer (m)

Boron (ppm)

Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

Control 28.5 3,800 6.15 592 5 784.1 48,318 34.04 23,264 15 536.3 33,426 25.35 15,355 30 528.9 24,515 19.14 11,278 50 188.8 21,439 18.47 9,986 75 161.4 15,696 18.29 5,316 100 130.2 13,191 12.33 4,018 Table 6 Concentrations of B, Cl, Li and Na in soils exposed during horizontal discharge; average of 9 collections Distance from the silencer (m)

Boron (ppm)

Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

Control 2.11 304.3 <0.10 111.7 5 35.39 1,914 2.33 958.0 15 17.01 1,457 0.56 501.4 30 9.67 794.3 0.47 311.8 50 7.75 553.6 0.18 219.1 75 6.74 529.7 0.21 198.8 100 7.55 450.1 0.16 185.7

10

Table 7 Concentrations of B, Cl, Li and Na in the leaves of natural vegetation during vertical discharge; average of 2 vertical discharges and 6 replicates/discharge Distance from the wellhead (m)

Boron (ppm)

Chloride (ppm)

Lithium (ppm)

Sodium (ppm)

Control 24 2,730 22 1,050 Pre-discharge 28 2,970 18 1,420 20 250 21,112 19 6,903 50 267 27,266 26 10,490 75 283 23,200 15 9,200 100 203 21,600 17 9,170 125 280 22,050 16 4,190 150 235 14,400 12 3,800 200 200 10,000 10 2,210 Table 8 Percent of test seedlings at different distances from the wellhead that have recovered from exposure to horizontal discharge; based on the average of 18 test seedlings per distance/station. 5 m 15 m 30 m 50 m 75 m 100 m 45% 58% 82% 92.5% not

affected not affected

Table 9 Tolerance limits of plants to salinity ions; after Cyberconference: water quality for woody plants, part I by Hannah Mathers of Oregon State University Salinity ions Upper Limits

(ppm) Optimum Range

(ppm) Boron 0.80 0.20 – 0.50 Chloride 140 0 – 50 Sodium 50 0 – 30

11

LUZON

0 300

MINDANAO

11°

121°

125°

19°

15°

Mt. Mayon

Mt. Apo

LGPF (Leyte)708 MWe

104 MWe

Mt. Canlaon

Mt. Bulusan

Taal Volcano

Mt. Pinatubo

PNOC-EDC Power projects

Volcanoes

Hibok-Hibok

VISAYAS

BATANES GROUP

150 MWe

192.5 MWe

km

NORTHERNNEGROS

Fig. 1. Location of the Northern Negros and other geothermal field in the Philippines

12

0 20 40 60 80 100 120 140 160 180 200Distance from well (m)

10

100

1000

10000

Chl

orid

e co

ncen

tratio

n (p

pm)

Leaves (horizontal discharge)

Soil

Spray catch

Washings

Leaves (vertical discharge)

1

Fig. 2. Variation in chloride concentration with distance from the well during the

discharge tests. Solid symbols are for measurements at sample sites, open symbols are for the control site (1.5 km away) for the horizontal discharges. Crosses and broken lines indicate chloride concentration on the leaves of natural vegetation for the vertical test. Note that the graph is log-normal. Instead of “spray catch” should it be “catch basins”?

13

A

B

C

Fig. 3. Test seedlings before horizontal discharge (A), one day after start of discharge (B), and after seven days (C)

14

A

B

C

Fig. 4. Natural vegetation before (A), nine days after the first vertical discharge (B), and three months after completion of the vertical discharges (C).

15