Best Combination? Rainwater Harvesting + Desalination

Note: Information provided are examples for discussion. Though it is based on a real and factual example, in the interest of privacy more details are withheld.

Background:
There is a development that needs to supply all of it's own water. It presently has a system that provides 220,000 litres of water per day, or equivalent to supply 1,000 people 220 litres of water per day. During the more rainy months they have a problem that the resort floods, producing problems in terms of loss of revenue.

The resort was looking to increase the water supply to 350 litres per person per day, this or what represents an increase of 130,000 litres per day. The first option that they had considered was to install another desalination plant or increase the desalination plant's capacity. We will analyse the costs involved in this. The options considered and why.

We were approached was because as there was abundant rain, so much so that there was flooding, that water harvesting would be an option.

The chart below shows the average rainfall pattern for this location.

The Different Options
As we were looking at different technologies, these mainly break down in three simple groups. An exclusive desalination option, an exclusive water harvesting option and a combined harvesting and desalination option. We will analyse each situation individually. I also provide a link to the spreadsheet where the calculations and graphs are obtained from.

Exclusive Desalination Option
We started with our analysis. The base information that we had of the initial expenditure for the desalination plant would permit us to extrapolate the further costs. The initial cost of the desalination was $1.2 million, so the increase of about 50% in capacity would imply an additional cost of  $600,000 approximately. The operating cost of the the plant was $5.5 per 1,000 litres. This would imply that apart from the Capital Expenditure for the plant to supply the complete requirement of 350,000 litres a day to make the system work would mean an Operating Expenditure of more the $700,000. This would be our base for comparison.

Exclusive Rainwater/Stormwater Harvesting Option
We were approached as we are know for the effectiveness and flexibility of our water harvesting option, which apart from being very flexible are also scalable, which allows them to be used in large civil infrastructure works as a viable and cost effective option.

Considering the rain data, even though there was a fairly abundant amount of water, more than 2,000 mm of rainfall a year, we immediately found that the specific demand of water would exceed even that amount of copious water, in addition to that, the catchment area placed serious limitations to what was possible. In any case the analysis had to be done. It was determined that the required catchment area would be in the vicinity of 160,000 square metres. To get an idea of what we are talking about in size, a football field has an area of about 10,000 square metres this implies, that we are talking about 16 full size football fields, which is a size that is not really available for catchment in this case. In addition to this it would need a rather large tank, a minimum size of 10,000,000 litres would be required. At an installed cost of $0.40 per litre, this would represent a total Capital Expenditure of $4,000,000. Though with this option the Operating Expenditure is reduced to practically nothing as all that is required are small pumps, to lift the water from the bottom of the tank at 1 metre of depth.

In any case it is interesting to make an analysis and see over a long period of time, if it were possible to have enough catchment area, what would be the savings. Below is the cost comparison that shows, the savings over a 10 year period.

Cost Analysis comparison of only Desalination vs. only Harvesting.

Project Cost Yearly Forecast
Only desalination			Only harvest
CAPEX	$600,000.00		CAPEX	$4,000,000.00
OPEX	$702,625.00		OPEX	$0.00
			OPEX Saving	$702,625.00
Current OPEX	$441,650.00		Install cost per litre	$0.40
OPEX Increase	$260,975.00		Recovery time (years)	4.84
10 year comparison
CAPEX + OPEX	$7,626,250.00		CAPEX + OPEX	$4,000,000.00
Total saving	$3,626,250.00
Saving Ratio	52.45%

Combined Harvesting and Desalination Option
The previous analysis is a bit of a futile exercise, even though it serves to show the long term benefits. What should be done is an analysis that is in fact feasible. The presence of the existing desalination plant is an advantage in terms of supply water during the drier months, in this case from July to November as the initial rain fall chart shows.

Whenever the storage tank falls bellow a certain level then the desalination units enters operation supply the necessary demand of water.

The model to be used follows the flow chart shown below.

Cost Analysis comparison of combined Harvesting with Desalination.

Project Cost Yearly Forecast Only desal Combined harvest + desal CAPEX $600,000.00 CAPEX $1,600,000.00 OPEX $702,625.00 OPEX $156,266.00 OPEX Saving $546,359.00 Current OPEX $441,650.00 Install cost per litre $0.40 OPEX Increase $260,975.00 Recovery time (years) 1.83 10 year comparison CAPEX + OPEX $7,626,250.00 CAPEX + OPEX $3,162,660.00 Total saving $4,463,590.00 Ratio 41.47%

The graph of how this system behaves and the amount of desalinated water that is provided is shown in the graph below. 

The spreadsheets, data and graphs for the above situation can be seen here:
https://spreadsheets.google.com/spreadsheet/pub?hl=en_GB&hl=en_GB&key=0ArEwIvQBUPQzdGRGZ01LNWo2UUI5TTMxLTRJYXpDdnc&output=html

Conclusions
From the above cost analysis it can easily be seen that the best solution is a combination of Rain/Stormwater Harvesting used together with a desalination unit order to provide water for the drier months. The higher capital expenditure that the installation of a large water storage tank implies is quickly recovered in less than 2 years and the saving over a period of 10 years are quite considerable amounting to just under $4.5 million, or $450 thousand per year. In addition to this as there is a large amount of water that is being captured and managed underground, the significant yearly flooding problems that the resort has would be averted.

Reiterative Reuse Water Harvesting Model

In my experience in working with water reuse one of the most common questions asked is how to appropriately size a tank system for water harvesting. Related to this there is a fair amount of information, and the costing is a simple consequence of the required volume. However paradoxically this modelling only considers only the one time use of the water, which is then "wasted" after being consumed.

A more efficient and cost effective possibility is available:

https://docs.google.com/drawings/pub?id=1hnz9I5flsTOT9TeTCunbNVLyjDROCHRdPCKB1sqpeTU&w=950&h=691

Recently I have been working on a rather large project, there is a "significant" requirement of water considering the source is just harvested rainwater and stormwater, that is 5 million litres of water a day. To propose a traditional system that considers only the one time use of water would not be feasible. For the economic and technical proposal a reiterative reuse of harvested water is necessary, and as such a new model had to be developed.

The above flow chart diagram shows a much more efficient model where water once consumed is reused reiteratively. The processing for reuse of the water use technology that is passive, practically does not require external energy and has very low maintenance, if any. The quality of the influent is one that can be considered as a grey water, and the resultant purified water quality is sufficiently high for all uses except potability, however this can also be achieved with relatively simple polishing of the water.

Source data webpage

The above graph shows the result of this modelling. However to arrive to a graph that has any meaning it is necessary to have to have previously done the relevant calculations. The spreadsheet for this can be seen  here: Reiterative Reuse Water Harvesting Model. This spreadsheet is modelled taking into account the following aspects.

Design Considerations
A prime and original consideration in terms of the modelling is made assuming the following situations:

Dual type tank system
The tank system is composed of 2 main types of tanks, a harvesting tank, that captures directly the rainfall, and a reuse tank, that has the double purpose of collecting harvested water as well the system that recaptures the water that has been consumed.

Reuse percentage
For practical effects the percentage of reuse is estimated at approximately about 80%, this can be defined to whatever is desired.

Tank Use Prioritization
The water from the reuse tank, for the effects of the model, is given priority in terms of consumption, once the capacity of the reuse tank is consumed, then the water from the harvest tank begins to be consumed. The prioritization of the reuse tank, is recommendable because it is desirable that that the water that is being reused is of less quality that the virgin harvested rainwater. Once water has received some degree of contamination it is recommendable that it be in movement and aerated as much as possible, in this case it receives this action in the course of its normal use. To leave contaminated water to sit a rest, will cause it to deteriorate in quality, entering a cycle of stagnation.

Modelling criteria
As the daily consumption is to be a considerable amount and maximum efficiency of use has to be obtained from the system. As such the modelling of the water level in the tank system has to be done on a daily basis. For this not only the data of the monthly rainfall has to be taken in consideration but also the frequency of that rainfall. When the system contains water to supply the full capacity that is required it will do so, however once the contents of the system are depleted, so as not to be able to provide the complete required capacity, what is available will be consumed, and also recycled in the percentage amount that has been defined, and will re-enter the system, making it available again for use the following day.

Modelling Resolution
As the modelling has a resolution to individually discrete days, the average frequency of rainy days is rounded up and then the amount of monthly rainfall is evenly distributed and averaged to evenly spaced days in the calendar month. The rounding up of frequency does not affect the total quantity of rainfall. This way an average quantity of rainfall that is captured per individual precipitation events can be estimated and then considered to be available for consumption.

Stabilization Period

The model is also performed for a period of 2 years to allow it stabilize. This is especially relevant in locations where rainfall is very seasonal, such as a monsoonal environment. It may however be the case when requirements are such that it is not possible to arrive to a stabilization, considering that complete provision of water is made available, such is what seen in 2 of the 3 examples of the modelling. In such case the recycled water will serve the purpose of being a complementary source. As such, as is normal available budget has much to do with what can be provided.

Controllable variables
Variables which are defined or controlled by the design engineer are the capacity of the tanks, as the water harvested or reused cannot physically exceed this volume.

Consumption defined
The consumption also can be defined by the design engineer. In relation to this to simplify the modelling a single daily consumption figure was chosen, however as the modelling has a daily resolution this figure can also be adjusted to increase consumption in periods of more abundance of rainfall and then reduced for periods of more scarcity. For demonstration purposes a suitable figure has been chosen that gives a balance of a reasonably high consumption, as much as possible while, while maintaining a minimum quantity of water in the system at all times.

Conclusions
The most interesting situation with this reiterative reuse model, that we have specifically studied, is that on a yearly basis that amount of water that is consumed through the recycling of harvested water is 3 times the amount of water that is used if the harvested water was consumed only once. This means it is possible to provide a client with a system which costs, practically a quarter, considering the initial use of water, of what the system would cost with out reuse.