Skip to main content

Full text of "How the feed temperature affects the cost of concentrating maple sap by reverse osmosis"

See other formats


Historic, archived document 



Do not assume content reflects current 
scientific knowledge, policies, or practices. 



SuSJ^/ ERRL Publ. No. 3730 



How The Feed Temperature 

Affects the Cost of 

Concentrating Maple Sap by Reverse Osmosis 

ARS-NE.4 
October 1972 



CO 



■ — S2 
m 




AGRICULTURAL RESEARCH SERVICE • U.S. DEPARTMENT OF AGRICULTURE 



This is a report of work done at the 



Eastern Regional Research Laboratory 
Agricultural Research Service 
U.S. DEPARTMENT OF AGRICULTURE 
600 East Mermaid Lane 
Philadelphia, Pa. 19118 



ABSTRACT 



It is economically advantageous to heat maple sap to 75** F. from its 
usual 35° to 45** storage temperature for processing by reverse osmo- 
sis. Experimentally, an increase in feed temperature from 55" to 74° 
produced more than a 30-percent average increase on membrane flux at 
pressures from 400 to 600 p.s.i.g. An economical heater is described 
for controlled heating of maple sap. 



3 



HOW THE FEED TEMPERATURE AFFECTS THE COST OF 
CONCENTRATING MAPLE SAP BY REVERSE OSMOSIS 



By J. C. Underwood and V. A. Turkot 
Eastern Regional Research Laboratory 
Philadelphia, Pa. 



INTRODUCTION 

The successful use of reverse osmosis as a means of separating pure water 
from a great variety ot dilute aqueous solutions has been due to the develop- 
ment of a semipermeable membrane with high water flux relative to that of the 
solute. The specially prepared cellulose acetate membranei./ is at present the 
only one generally recognized as practical for reverse osmosis applications. 
In the work, to develop the cellulose acetate membrane with high water flux, 
the effect of temperature on membrane permeability was recognized. An increase 
in the environmental temperature of the membrane increased its permeability^/. 
However it has been found also that membrane deterioration is faster at higher 
temperatures. A maximum temperature of 80° F. for extended usage has been 
recommended by manufacturers of commercial membrane units. 

In the maple sirup industry the raw maple sap is stored at low temperature 
(35 to 45° F.) until processed to sirup by atmospheric boiling. There are no 
specific data in the literature on the effect of feed temperature on the water 
flux of the membranes in a commercial reverse osmosis unit. In the case of 



i_/Loeb , S~. Preparation and performance of high-flux cellulose acetate desali- 
nation membranes. _In "Desalination by reverse osmosis," Ed. by U. Merton, pp. 
55-91. The MIT Press, Cambridge, Mass. 1966. 

eid, C. E. , and Kuppers, J. R. Physical characteristics of osmotic mem- 
branes of organic polymers. J. Appl . Polymer Sci. 2: 264-272. 1959. 

Loeb, S., and Sourirajan, S. Sea water demineralization by means of an os- 
motic membrane. Advan. Chem. Ser. 38: 117-132. 1962. 

Wiley, A. J., Ammerlaan, A. C. F. , and Dubey, G. A. Application of reverse 
osmosis to processing of spent liquors from the pulp and paper industry. Tappi 
50: 455-460. 1967. 



4 



maple sap, the increase in water removed from the sap at a higher feed temper- 
ature could be the deciding factor in determining the economic feasibility of 
using reverse osmosis as part of the sirup-making procedure. This paper re- 
ports the effect of feed temperature on the permeability of the cellulose 
acetate membrane used in a pilot reverse osmosis unit of 10,000 gallons a day 
feed capacity. 

EXPERIMENTAL 

Materials and Methods 

A pilot reverse osmosis unit— ^ designed and built for commercially field- 
testing maple sap concentration^' was used in this study. Six hundred square 
feet of cellulose acetate osmotic membrane in the form of the rolled module 
were utilized, three modules (each 3 feet long) in each of four 10-foot pres- 
sure tubes. The feed being used, either tap water or dilute sugar water, was 
pumped through the pressure tubes at AOO, 500, and 600 p.s.i.g. at a rate of 
6 gallons per minute. The feed was stored in a 200-gallon holding tank 
equipped with a jacket through which refrigerated water was run to maintain 
desired temperatures. Both discharges from the unit, permeate and concentra- 
ted feed, were returned to the holding tank to minimize the volume of liquid 
needed for the study. The rate in gallons per minute, temperature, and con- 
ductivity of the feed were recorded after a 10-minute run at 400, 500, and 
600 p.s.i.g. in the pressure tubes. The temperature of the feed was measured 
in the feedline between the pump and the pressure tubes. This gave a value 
equivalent to the temperature of the liquid in the pressure tubes. No temper- 
ature change occurred in the feed during its passage by the membranes in the 
pressure tubes after an equilibrium with tube temperature was reached. The 
feed temperature was controlled by the cooling water in the holding tank. 

The rates of feed, concentrate and permeate, were obtained from meters 
on the unit whose accuracy had been verified by volume per time measurements 
of the discharges from the unit. The conductivity of the permeate was ob- 
served throughout the study as an indication that the unit was working properly. 

Tapwater was used as feed to avoid the inaccuracies of pressure values 
caused by unmeasurable changes in osmotic pressures. The osmotic pressure of 
a feed containing a solute increases as water is removed from it. Flux (water 
passing through the membrane) is directly proportional to the working pressure 
of the system, the applied pressure less the osmotic pressure of the feed. 



_' Moore, W. H. , Jr., and Willits, C. 0. A reverse osmosis plant for maple 
sap concentration. U.S. Dept. Agr. , Agr. Res. Serv. , ARS 73-66, 13 pp. 1970. 

— Underwood, J. C. , and Willits, C. 0. Operation of a reverse osmosis plant 
for the partial concentration of maple sap. Food Technol. 23: 787-790. 
1969. 



5 



Procedure 



Two hundred gallons of tapwater were run into the holding tank and cooled 
with stirring to a preselected temperature by the action of the cooling water 
in the tank jacket. Then the reverse osmosis unit was started and the water 
in the tank pumped through the unit at 400 p.s.i.g. membrane pressure until an 
equilibrium was reached near the preselected temperature. Then the temperature 
in the feed line was recorded, along with the gage readings for feed and con- 
centrate flow rate. After 10 minutes, a second set of data was taken. The 
pressure on the membranes was then raised to 500 p.s.i.g. After a 10-minute 
equilibrium interval, data was again recorded. This was repeated for 600 
p.s.i.g. 

The temperature of the coolant in the jacket of the holding tank was then 
raised to produce a higher temperature feed. Data was recorded for this 
temperature at 400, 500, and 600 p.s.i.g. membrane pressures. In such fashion 
data was obtained for feed temperatures of 55"*, 62°, and 74° F. 

RESULTS AND DISCUSSION 

Recorded in table 1 are the water fluxes for the cellulose acetate mem- 
branes in the pilot unit for feed temperatures of 55°, 62°, and 74° F. and 
applied membrane pressures of 400, 500 and 600 p.s.i.g. These data show the 

TABLE 1. — Effect of feed temperature on the flux of the cellulose 
acetate reverse osmosis membrane 



Pressure 


Feed temperature, °F. 


55° 


62° 


74° 


P.s.i.g. 


Gallonsiy 


Gallonsiy 


Gallons!/ 


400 


6 


7 


8 


500 


7 


8 


9 


600 


8 


9 


11 



Gallons per day per square foot membrane. 



very significant effect of temperature on the permeability of cellulose ace- 
tate membrane. At the three pressures tested, more than a 30 percent average 
increase in flux was obtained by raising the temperature of the feed from 55° 
to 74° F. At 600 pounds applied pressure, the increase was 37.5 percent. 
This value can be used to measure the increase in capacity of the unit at the 
higher temperature. Increased capacity would mean reduced capital investment. 



6 



As the storage temperature of maple sap is usually below 50° F., the advantage 
of heating the feed to a reverse osmosis unit would be greater than indicated 
by the values reported in this study. 

To illustrate the economic advantage of heating the maple sap feed for 
reverse osmosis concentration the following example is presented. 

In this example, the following conditions are assumed. The reverse osmo- 
sis unit (R.O. unit) will be used as a preconcentrator. It will be fed with 
raw sap, and its output will go to conventional evaporating equipment for fin- 
ishing to maple sirup. For calculation purposes, the solids content of the 
raw sap is assumed to average 2.5 percent and the solids content of the R.O. 
output to average 10 percent. These figures represent removal by the R.O. 
unit of just over 75 percent of the water present in the raw sap. Loss of 
solids in the R.O. unit is very slight and can be ignored for these calculations. 

The sugar house where the R.O. unit is installed is assumed to produce a 
total of 5,000 gallons of sirup during the season. This will require a sea- 
sonal total input of raw sap, at 2.5 percent solids, of about 172,000 gallons. 
It is also assumed that the season will include 24 days of actual sirup-making 
operations, with an average of 12 hours of operation each day. This gives a 
total for the season of 24 x 12 = 288 operating hours for the R.O. unit. The 
hourly rate of sap being processed will thus be 172,000 gallons divided by 
288 hours, or about 597 gallons per hour. 

In operation, the sap will be drawn from a storage tank, passed through a 
heater, and then fed to the reverse osmosis unit. The sap is assumed to enter 
the heater at about 45" F. and be warmed to about 75* F. It is assumed that a 
steam boiler is already being used by the sugar house, and that it will have 
enough capacity above its normal load to supply the sap heater with 150 pounds 
of steam an hour (4 1/2 boiler horsepower). 

The R.O. unit will be operated with a feed pressure of 600 p.s.i.g. and a 
feed rate of about 597 gallons per hour. To deliver product at 10" Brix, the 
R.O. unit must separate out about 450 gallons of water (permeate) from the sap 
per hour. If the unit were fed with unheated sap, that is, at a temperature 
of about 45° F. , its permeate rate would be about 6 gallons per 24 hours per 
square foot of membrane, or 0.25 gallons per hour per square foot. Thus, the 
total membrane area required would be 450 divided by 0.25 = 1,800 square feet. 
When the unit is fed with preheated sap at 75°, however, its permeate rate 
will be about 9 gallons per day per square foot or 0.375 gallons per hour per 
square foot. The total membrane area required will then be 450 divided by 
0.375 = 1,200 square feet. Thus, preheating the sap to 75° reduces the re- 
quired membrane area by one-third. The difference in factory cost for the 
two different size R.O. units, one with 1,800 square feet and the other with 
1,200 square feet, would be at least the saving in membrane cost, about $5.00 
per square foot or $3,000. This saving would be carried forward to the re- 
placement of membranes which cost at present $4.00 per square foot. 

This capital savings would have to be reduced by the cost of the heater 
plus the energy cost of raising the temperature of the permeate (water 



7 




Q 

LU O 
LU O 







M 








o 








4-1 








o 








E 




CSJ 






• 


CN 


> 


T3 




t 


1— ( 


C 


3 


o 




CO 


W 


CO 


> 




T3 










** 




£ 


>-i 


QJ 


tH 


3 


O 




O 


D> 


U-) 


3 


C 








QJ 


f— 1 






rH 


CO 


CO 


Vj 


O 


00 


CO 


0) 


(0 


3 




o. 




cw 




e 


l-l 


•H 


c 


(U 


o 


V4 


(U 


4-) 






'O 






c 


o 


a. >H 


0) 


o 


n) 


o 


o 




(0 


t-l 








4J 


^ — V 




"O 








0) 


o 






4J 


u 






(0 








31 


4-1 


w 




4= 


tfl 


0) 


T— 1 




4J 


4J 


cd 


60 


10 


CO 


4^ 


c 


(B 




o 


•H 


3 




*^ 




cr 


00 




o 


CO 


c 


4J 






"H 


<L) 


(0 


>-i 


X 


<D 




o 


•H 




u 




E 




(U 


4-1 




o 


4-1 


Cfl 




-3- 


0) 


4-1 


c 


r-l 


e 


CO 


CO 


V — ' 


o 


O 






e 


e 


00 o 


u 


u 


c 


1 


01 


(U 


•H 


ro 


J= 


j: 


4-1 


O 


u 


4J 


CO 


O 






0) 




\ 




4= 




CM 


•* 




o> 




N ' 


u 


a 






o 










4J 


d 








o 


b 


V4 




4-) 




o 


60 




o 


4J 


C 


c 


O 


CJ 


•H 




CN 


0) 


ja 


D. 


•r-) 


3 


O 


1 


c 


4-1 




o 


•H 






m 






4J 








•H 






c 




e 




•H 




3 


4-1 


E 


e 




CO 


3 


D 


•a 




iH 








CO 


■a 


c 


vO 






•H 






i-t 






Q 


<u 


V4 






(U 


0) 


>4-l 


d 


4-1 


4J 


14-J 




CO 


CO 


o 






3 




o 


c 




•V 


c 


o 


tw 


c 


Ti 


iH 


O 


CO 


1 


rH 








(d 


01 


c 




60 


u 


o 


cn 


1 


3 






m 


4J 


e 


>4-l 


m 


CO 


CO 


o 




u 


o; 






0) 


4-1 






o. 


(0 


x: 




s 




CJ 




0) 




CO 




4J 


e 


OJ 




00 4J 


CO 



0) c 

4.1 i-( 

CO 3 
01 o 

00 

o. 

CO u 
CO 0) 



c 
o u 
u 3 

4-t 

> m 

H rH 
CO 

> •> 



< 0) 
I B 
I o 

0) 

0) £ 
»J 4J 

3 

60 /-N 

iH ro 
b — ' 



separated out by the R.O. unit). This energy cost, based on fuel oil for heat- 
ing, is estimated at about $55 a season. The heat energy put into the "con- 
centrate" (10-percent sugar product) from the R.O. unit is not lost, since this 
stream must be raised to the boiling point by some means. 

As the liquid passing through the R.O. unit should not be warmer than 
80° F., a heater with some controlled operation should be used. One such unit 
described in the following section can be built for about $250. Thus, there 
would be a definite economic advantage to heating maple sap to about 75° for 
processing by reverse osmosis. 



Design of a Sap Heater 

A number of different methods could conceivably be employed for heating 
the sap ahead of the R.O. unit. However, the method judged as best by the 
authors makes use of hot water as the medium to transfer heat to the sap. 
This method has the following advantages: relative freedom from operating 
difficulties which might necessitate shutdown of the heater; ease of control; 
minimum danger of overheating the sap which could cause damage to the expen- 
sive R.O. membranes; and freedom from rapid swings in temperature. Steam 
(assumed to be available) is used to heat the hot water, which in turn heats 
the sap. 

One version of such a sap heater, which could readily be built by anyone 
with mechanical apitude, is shown in figure 1. It employs a 55-gallon, open- 
top steel drum which, in operation, is nearly filled with hot water. The 
water is kept hot by direct injection of steam. The steam also agitates the 
water so as to keep it mixed and at a uniform temperature. Two coils of 
aluminum tubing are immersed in the hot water. The sap passes in parallel 
through these coils at a constant total rate of 10 gallons per minute, and is 
heated to 75° F. Temperature of the hot water is rea,d on dial thermometer 
(Item 3) and that of the heated sap on thermometer (Item 2) . Temperature of 
the sap is controlled by adjusting the temperature of the hot water. The 
latter is controlled by turning the steam flow on and off at electrically- 
operated solenoid valve (Item 5) . This valve is operated by automatic switch 
(Item 4), the control setting of which is adjustable over a range of tempera- 
ture. Item 4 is an ordinary aquastat type switch used on some domestic hot 
water heating systems, and operates (as does valve. Item 5) on regular 120- 
volt A.C. power. The temperature of the water in the heater, with cold sap 
entering at 45° F. , would normally be held at about 130° F. 

The drop in pressure of the sap while flowing through the heater would be 
about 8 pounds per square inch, or equal to about an 18-foot gravity head. 
If sufficient gravity head is not available, a small pump should be installed 
ahead of the heater to supply this pressure. 

Cost of the materials needed to construct the heater would be under $200. 
Cost of a pump and motor for the sap, if needed, would be about $65. Thus, 
total materials costs for the heater and pump would be about $260 (see table 
2). This figure does not include any charge for labor. 



9 



TABLE 2. — Cost of pump and heater 



Estimated 



cost 

55-Gallon steel drum (second-hand) 3.00 

Thermometers (2) 30.00 

Pipe fittings, valves 30.00 

Steam injector 20.00 

Pump with motor 65.00 
Controls: 

Thermostat (aquastat) switch 15.00 

Valve (solenoid) for steam 50.00 

Aluminum tubing (3003-0) 35.00 

Tubing fittings 3.00 

Wire, switch for aquastat and solenoid 6.00 

valve — 

Total $257.00 



If steam is not available for the sap heater, a package type hot water 
heater, either gas-fired or oil-fired, could be used. The capacity of this 
heater, expressed in several ways all equivalent to each other, would have to 
equal (or exceed) the following rating: 

1. 150,000 BTU net output per hour. 

2. 215,000 BTU input per hour (70-percent efficiency). 

3. Recovery (heating) rate of 180 gallons of water per hour at 
100° F. temperature rise. 

4. Recovery (heating) rate of 300 gallons of water per hour at 
60* F. temperature rise. 

A hot water heater of this capacity, designed to operate on L-P gas, 
would cost about $450. A pump to circulate the hot water should also be 
installed. This would cost about $60. 



10 



CONCLUSION 



This study shows that there is an economic advantage to passing maple sap 
through a reverse osmosis unit at a temperature higher than the usual storage 
temperature for this raw material for maple sirup. However, the cellulose 
acetate membrane used in reverse osmosis is more stable at lower temperatures. 
Therefore, when all factors are considered, a feed temperature near 75° F. 
is recommended for reverse osmosis concentration of maple sap. 

Since the temperature of the heated maple sap feed to the R.O. unit 
should not be above 80° F. beyond a brief period, a heater with controlled 
operation should be used. A unit has been described for this purpose employ- 
ing hot water as the heating medium. This unit can be built for about $260. 
However, many other ways of heating the sap are possible and the chief re- 
striction on any method would be the necessity for some control of the heating. 



11 



UNITED STATES DEPARTMENT OF AGRICULTURE 
AGRICULTURAL RESEARCH SERVICE 
EASTERN REGIONAL RESEARCH LABORATORY POSTAGE AND FEES PAID 

600 EAST MERMAID LANE U g. DEPARTMENT OF AGRICULTURE 

PHILADELPHIA, PENNSYLVANIA 19118 AGr 101 



OFFICIAL BUSINESS 

PENALTY FOR PRIVATE USE, 



$300