AGRISOL Project - Preliminary Data

This section includes a schematic description of the various components that are currently installed in the NF pilot desalination plant in Hatzeva and an overview of the experimental results obtained from operation of the pilot unit with grid electricity.

System details

Membranes

Membrane elements: Two Dow Filmtec NF90-4040

Salt rejection: 99.2% (99.0% minimum)

Configuration: Spiral wound

Polymer: Composite polyamide

Total membrane active area: 15.8 m²

HP feed pump

Pump model: Grundfos SQFlex 1.2-2

Maximum current: 8.4 A

Supply voltage: 30-300 VDC, PE

1 x 90-240 V -10%/+6%, 50/60 Hz, PE

Maximum flow rate: 1.4 m³/hour

Maximum pressure head: 120 m

Renewable energy system

Maximum power: 4 kW_p(15 x 280W)

Battery capacity: 12 rechargeable batteries, 2V 750 Ah

Control unit: 2.1 kWh inverter to provide constant V to the pump

MPPT: TriStar MPPT 60 Amp

Table B1 shows the flows and pressures measured in the pilot desalination unit between November 2010 and February 2011. During this period, the NF pilot desalination unit has achieved the objectives and exceeded the expectations for faultless operation, permeate production and energy consumption. The NF pilot has been in operation for 24 h/d over the whole period and without any interruption due to defective operation.

Table 8. Operation parameters of NF desalination unit between November and February

Date		Flows				Pressures
	Permeate	Recycle	Brine	Blending	Feed	Before NF	Brine
	L/h	L/h	L/h	L/h	bar	bar	bar
29/11/10	270	900	200	25	2.0	4.2	4.0
2/12/10	280	900	200	25	2.2	4.3	4.0
6/12/10	270	800	200	25	2.0	4.4	4.1
9/12/10	280	900	200	25	2.3	4.2	4.2
12/12/10	260	1000	200	25	2.0	4.3	4.1
16/12/10	260	900	200	25	1.8	4.3	3.9
20/12/10	270	800	200	25	1.8	4.2	4.1
24/12/10	270	800	220	25	1.5	4.3	4.1
27/12/10	270	800	200	25	1.8	4.2	4.0
30/12/10	260	900	220	25	1.6	4.3	4.2
3/1/11	270	900	200	25	1.7	4.2	4.0
5/1/11	270	900	200	25	1.6	4.2	4.0
10/1/11	270	900	220	50	1.7	4.3	4.3
13/1/11	280	1000	220	50	1.8	4.4	4.2
17/1/11	270	900	200	50	1.8	4.3	4.2
20/1/11	260	900	200	50	1.7	4.2	4.2
24/1/11	270	1000	220	50	1.6	4.4	4
27/1/11	280	900	220	50	1.8	4.3	4.1
31/1/11	260	900	210	50	1.7	4.3	4.2
3/2/11	280	800	200	50	1.6	4.4	4.1
7/2/11	270	900	220	50	1.8	4.2	4.1
10/2/11	280	800	200	50	2.0	4.3	4.2
14/2/11	280	900	220	50	1.7	4.4	4.1
17/2/11	270	900	220	50	1.8	4.2	3.9
21/2/11	280	800	200	50	1.6	4.3	4.0
24/2/11	280	900	220	50	1.7	4.3	3.9
28/2/11	270	900	220	50	1.7	4.2	4.1

The NF pilot consistently produced a permeate flow of 271 L/h (± 7 L/h) during the period of observation, corresponding to a daily production of 6.5 m³. This is notably higher than what expected when the system was installed (5 m³/d) and mainly due to a higher feed inflow (480 L/h). To ensure smooth operation of the NF membranes, the recovery rate was maintained at a conservative level (57%). Chemicals to prevent scaling for pH control were nevertheless continuously added. Due to the higher feed flow and conservative recovery rate, the brine flow was higher than in the design scheme (209±10 L/h ). The operating pressure of the NF membranes was stable in the range 4.2-4.4 bar and did not show any increase over the period of operation. The blending flow was initially set at 25 L/h and subsequently raised to 50 L/h on January 19 in order to increase the concentration of micronutrients (Ca²⁺ and Mg²⁺) in the permeate flow. The specific energy consumption of the pilot plant was 1.37±0.02 kWh/m³. The pilot desalination plant worked with grid electricity during this period of operation.

Samples of feed water, permeate, and brine were collected on a weekly basis and analyzed in the laboratories of the Ben-Gurion University of the Negev for the concentration of the ions of principal interest to this study (Ca²⁺, Mg²⁺, SO4^2-, Cl^-, HCO3^-), pH and electro-conductivity (EC). In addition to these measurements, the concentration of K+ and NO3- was measured on a non-regular basis. Table B2-B4 show the results of the chemical analyses.

Table 9. Chemical composition of feed water between November and March

	Ca²⁺	Mg²⁺	SO₄^2-	Cl^-	HCO₃^-	K⁺	NO₃^-
	ppm	ppm	ppm	ppm	ppm	ppm	ppm
29/11/10	190	90	560	350	235
6/12/10	185	90	555	320	215
12/12/10	185	90	560	330	215
19/12/10	185	90	560	325	215
27/12/10	170	82	409	277	254	9.6
6/1/11	208	95	526	331	245	10.3
16/1/11	215	100	547	376	230	10.6	33.8
30/1/11	170	81	490	321	232
7/2/11	186	91	513	427	280
13/2/11	165	78	487	302	250
27/2/11	162	85	550	360	215
6/3/11	151	79	516	360	210
Average	181	88	523	340	233	10.2	33.8

Table 10. Chemical composition of permeate between November and March

	Ca²⁺	Mg²⁺	SO₄^2-	Cl^-	HCO₃^-	K⁺	NO₃^-
	ppm	ppm	ppm	ppm	ppm	ppm	ppm
29/11/10	28	14	85	120	52
6/12/10	25	13	80	110	45
12/12/10	27	14	85	115	46
19/12/10	26	14	80	115	47
27/12/10	17	8	35	78	37	3.2
6/1/11	21	9	43	93	41	3.8
16/1/11	20	9	43	93	35	3.3	20.1
30/1/11	44	21	120	126	75
7/2/11	53	27	146	162	66
13/2/11	36	18	110	102	67
21/2/11	44	21	135	130	72
27/2/11	50	27	172	164	77
6/3/11	43	24	150	154	73
Average	33	17	99	120	56	3.4	20.1

Table 11. Chemical composition of brine between November and March

	Ca²⁺	Mg²⁺	SO₄^2-	Cl^-	HCO₃^-	K⁺	NO₃^-
	ppm	ppm	ppm	ppm	ppm	ppm	ppm
6/12/10	430	210	1290	780	490
12/12/10	455	215	1320	810	495
19/12/10	460	220	1360	780	525
27/12/10	410	193	1054	619	532	20.4
6/1/11	425	200	1110	654	493	19.1
16/1/11	498	230	1347	778	524	20.1	49.1
30/1/11	426	203	1220	714	546
7/2/11	462	231	1280	860	450
13/2/11	400	192	1216	622	567
27/2/11	390	208	1420	760	486
6/3/11	384	208	1330	780	500
Average	431	210	1268	742	510	19.9	49.1

The chemical analyses of the feed water show that little variation occurred during the period of observation in the inflow concentrations of the monitored ions with the exception of Ca²⁺, which shows a decreasing trend from the first measurement in November (190 ppm) to the latest measurement in March (151 ppm) that is statistically significant at the 10% level (pvalue = 0.07). A corresponding decreasing Ca²⁺ concentration is observed in the brine.

The desalination unit has performed according to model forecasts regarding the removal of the monitored ions. Retention of Cl^- ranged between 54% (February 27) and 75% (January 16), corresponding to an average Cl^- concentration in the permeate of 120±26 ppm. With the initial blending flow set at 25 L/h, the concentrations of Ca²⁺ and Mg²⁺ in the permeate were, respectively, 23 ppm and 12 ppm, values which are lower than the recommended minimum concentration of these nutrients for agricultural use. Consequently the blending flow was increased on January 19 to 50 L/h, raising the average concentration of Ca²⁺ and Mg²⁺ in the permeate to 45 ppm and 23 ppm, respectively. Such values are well within or even above the recommended range of concentrations (32-48 ppm for Ca²⁺; 12-18 ppm for Mg²⁺). The concentration of SO4^2- was well above the minimum recommended value of 30 ppm during the whole period.

The EC of the feed water shows little variation over the whole period, averaging 2.32 dS/m (±0.15 dS/m) and with no trend towards increase or decrease over time. Similarly, the EC of the brine is relatively constant at 4.73 dS/m (±0.25 dS/m) as it is unaffected by the change in the blending rate on January 19. The EC of the permeate flow averaged 0.71 dS/m (±0.18 dS/m) over the whole period of measurements. Before the increase in blending flow, the average EC of the permeate is 0.58 dS/m (±0.09 dS/m); after January 19 the average EC increases to 0.86 dS/m (±0.12 dS/m), a value that is satisfactory since it is still within the range tolerated by the most salt-sensitive crops.