Dear all,

I recently had an interesting conversation on Twitter with Catarina Fonseca (she is a worldclass expert on WASH costs and financing, working at IRC in the Netherlands). I thought I should copy it here in a way of archiving and because not everyone uses Twitter. Excuse the brevity of the questions and answers but that's the nature of Twitter:

Starting question:

Elisabeth von Muench ‏(@Ostella42) on 17 June:
Does 0.5 cent per user per day sound like the right cost for a sanitation service? @GatesFoundation @FonsecaCatarina

Replies:

Catarina Fonseca (‏@FonsecaCatarina) on 17 June
That's $182.5 per person per year, if all life-cycle costs inc. does fit into benchmarks www.washcost.info/page/2439

Elisabeth von Muench:
Thanks. Although isn't this rather on the high side? I found the benchmarks paper a bit confusing.

Catarina Fonseca:
Based on the 4 countries and if you take into account all costs... not really... let me know where I can help..

Plus it really depends on what you are trying to provide - a sanitation service is different from a latrine

Elisabeth von Muench:
Looking at it from the other angle: what cost is affordable (to be < 2.5% of household income)?

Catarina Fonseca:
That's a whole different question. There is no evidence or even %benchmark for what is affordable at HH level

But my PhD will be about that - so in 6 months we might have some answers.

Catarina's Twitter photo:


Regards,
Elisabeth

Hi Elisabeth,

good to point to the washcost project, this is very relevant for this discussion here.

Here an excerpt from the washcost material showing what cost elements should actually be considered when talking about full life-cycle costs or operational costs, most of them are actually always never considered in cost data:



Copied from this brochure here: www.washcost.info/content/download/1152/...ASHCost_Brochure.pdf

Regards, Florian

Dear Florian,

Did you find anywhere in the various publications from the WASHCost project an answer to Christoph's questions above from 6 June? I.e. what values to use for life time (in years) and discount rate (in %) to convert a capital cost into an annualised cost for sanitation infrastructure and services? I haven't found it. It must be in there somewhere. I regularly feel a bit frustrated with those WASTCost publications. They are nice, but somehow not practical enough, for my liking.

So that 1-pager that Jonathan suggested above, is something that I would very much support.

Regards,
Elisabeth

muench wrote: I.e. what values to use for life time (in years) and discount rate (in %) to convert a capital cost into an annualised cost for sanitation infrastructure and services?

Dear Elisabeth,

I don't know all the washcost publications in detail, but I think that there can't be any general valid values of life time and discount rate.

I am rather a layman when it comes to economics, but I understand that capital costs are composed of the two fractions described in the graph above: capital expenditure and cost of capital. If you take a commercial or development bank loan for financing your infrastructure, the annualised capital costs are what you pay back to the bank every year, that is a part to pay back the actual loan (capital expenditure) and the interests (cost of capital). So the annual costs are determined by the loan conditions and the time you choose to pay back the loan (which should ideally be equal or shorter than the designed life-time of the infrastructure).

If the infrastucture is financed by a grant or partly subsidised, I would take the designed life-time of the infrastructure (that is the time until the new investments are needed to replace or rehabilitate the infrastructure) and local market conditions for the interest rates. That is basically what Christoph suggests in his examples.

I don't think standardised values for life-time and interest rates can be used, as this depends much on the technology, local conditions and other aspects. What is important is that the calculations and the values (e.g. life time) are made transparent, so one can assess if they are realistic.

Best, Florian

Dear Florian,

Thanks a lot for your - as always! - helpful post.
I agree that one cannot give general values that hold true in all cases. But I am referring specifically to the value stipulated by the Bill & Melinda Gates Foundation's Reinvent the Toilet Challenge. Don't you think that if they give a value (0.05 US-cent per user per day), then there should also be agreed values on:

• Life-time
• Discount rate
• Maximum allowable number of users per toilet?

Because maybe otherwise, every grantee can just pick and choose values that make his or her technical solution look good, i.e. very affordable?
Or would it at least be possible to determine typical values on a country basis. E.g. use discount rate x in Bangladesh but discount rate y in Kenya.

Or do I have a view that is too simplistic? (Like you, I am a layman - or rather laywoman - with respect to economics)

Regards,
Elisabeth

muench wrote: But I am referring specifically to the value stipulated by the Bill & Melinda Gates Foundation's Reinvent the Toilet Challenge. Don't you think that if they give a value (0.05 US-cent per user per day), then there should also be agreed values on:

• Life-time
• Discount rate
• Maximum allowable number of users per toilet?

Hi Elisabeth,
I am not so sure what to think about this criteria. I feel that fixing a life-time and max. number of users is restricive would necessarely be not approriate for many situations. But then, already the 5 c criteria is kind of arbirtary, and as far as I remember there are other quite restrivie and not very realistic criteria in that call.

The good thing is that the 5c criteria pushes people to seriously think about costs. Of course tweeking of calculations may happen. I think best would be to require cost calculations that include the different cost components and justification of assumptions, so that the calculations can be easly assessed.

Regards, Florian

Hello all,

I have placed together my thoughts on a short document outlining (what I believe to be is a) consistent approach to costing sanitation systems. There is a lot of discussion on this forum already about costing systems / approaches and it is not my intention to have those discussions again.

This document is intended to outline an approach for specific units or technology types (hardware). It may be appropriate for people working in developing countries under a funded project etc. who would like to provide an estimate of costs if the system was to be applied in a developing region. It does not take into account the significant ‘software’ costs associated with providing a sustainable sanitation service. The IRC is undertaking this extensive package of work through the WASHCost project.

I have attached this document to encourage discussion and I take no offence if it is pulled apart of disregarded completely. I think the important outcome is the development of a consistence and transparent costing approach that can be easily understood and disseminated.

Chris

Dear Elisabeth and Florian,
that was exactly my point when I made the post. It is very easy to “tweak” calcs, therefore it is crucial to put in the basic assumptions. Just an number (cent/use) is not enough.
I do agree totally with Florian, it is not possible to give standardized lifetimes. Even with the same material, lifetime might differ due to the surrounding conditions.
As well it is not possible to give a general real interest or discount rate, that depends very much on the conditions in every country . Again therefore it is important to know which rate has been used.
And conclusions might be different in extreme cases, that makes the demonstration of the basic data so important, than everybody might just change the settings to their data.
Yours
Christoph

Dear all

Since I somehow started this post, I think it is time to explain how our RTTC team from Eawag/EOOS( www.diversionsanitation.com ) have dealt with the RTTC call requirements in a business case.

We have calculated an initial and preliminary business case for our “diversion for safe sanitation” system (consisting of blue diversion urine-diverting toilets and resourece recovery plants (RRP)). The context for the business case is Kampala, Uganda. We assume that a private firm would somehow start our proposed sanitation business. Initially, the private firm promotes rental agreements for our blue diversion urine-diverting toilets. Once, the private firm has reached a certain size in terms urine volume, it starts to produce and promote urine-based fertilizer for the market.


We used the following numbers for the business case:

Revenues

• Toilet rental fee: 0.05$/person/day (as stipulated in RTTC call)
• Users per toilet: 10 (about 2 families sharing one toilet)
• Expected revenues from fertilizer sales: 0.02$/person/day

Costs

• Toilet life span: 10 years (= life span of the membrane in our toilet)
• Toilet investment costs: 500$ (target costs, two toilets necessary throughout the life span of the business).
• Toilet operational costs: 10% of toilet investment costs (experience from industry)
• RRP investment costs: 27’500$ (target costs)
• RRP operational costs: 15% of RRP investment costs (experience from industry)
• Depreciation costs for toilets: 10% of the investment costs (= yearly saving to guarantee necessary reinvestments into new toilets after 10 years)
• Depreciation costs for RRPs: 5% of the investment costs (= yearly saving to guarantee necessary reinvestments into new RRPs after 20 years)
• Logistic costs: 0.01$/person/day
• Scale: 860 toilets are connected to one RRP 1 (based on logistic modeling)

Life span of business: 20 years (as stipulated in RTTC call)

Final scale of the business was operating 20 RRPs, i.e. renting out 1720 toilets which are used by 17200 slum residents.

We calculated the return on the employed capital per year (ROCE) as 7.96%. The ROCE value of 7.96% shows that the diversion business model achieves an interest rate of 7.96% on the capital employed per year. This above the recommended cost of capital of 6% per year which were stipulated in the RTTC call. Meaning, we showed that a viable sanitation business can be possible.

Next steps
We are aware that the business model still has a couple of drawbacks. We will start working on the drawbacks and find different versions of a business model.

Questions & comments are welcome.
If we can assist your team in any way on building such a business case and calculating revenues, costs, and profit, please do feel free to contact us (This email address is being protected from spambots. You need JavaScript enabled to view it. and This email address is being protected from spambots. You need JavaScript enabled to view it.).

Regards
Ulrike

Dear all,

I would like to come back to this discussion on costs (thanks Ulrike for your useful post above). My main focus in this post is the annualised capital cost, because that one is easier for the layperson to grasp.

In the thread above, Jonathan had proposed a 1-pager so that we are all “on the same page” and speak the same language.

Chris Sullivan has made a first draft, which I am very grateful about (please scroll up to 26 June to see it). I think it reads well and is useful. However, it is a bit too long (2 pages) and page 2 goes into detail for calculating costs to obtain the net present value (NPV) or life cycle costs. Whilst this is the more accurate way of doing it (and is quite easy to set up in Excel), it is not easily understood by the layperson, as you end up with a large number that doesn’t mean anything to a layperson.

(For these more accurate calculations on NPV, I very much look forward to having the WASHCost Tool available by IRC, which is right now being tested and launched at the World Water Week. See separate discussion on that one here:
forum.susana.org/forum/categories/55-wg-...cost-calculator#5476 )

I propose here to focus on the opposite, i.e. take the large number (capital cost) and convert it into an annualised or even daily cost. This is what the Gates Foundation did in their Reinvent the Toilet Challenge (RTTC).

Already in their first sentence for their request for proposals in the RTTC Round 1 challenge in February 2011, they included this famous 5 Cent figure:

The Bill & Melinda Gates Foundation announces a new challenge aimed at concept development, design, and prototyping of a means of dealing effectively and cost-efficiently (less than $0.05 per person per day) with human waste…

You can read the original request for proposals from Feb. 2011 here: www.susana.org/lang-en/library/library?v...eitem&type=2&id=1636 (first document of four which are available for download in that library entry).

On page 7 in that document it also says:

Because this challenge is at its essence an extreme value-engineering one, grantees are asked to make an ad hoc technical-economic argument to consider the likely ability of the proposed effort to ultimately lead to the capability to satisfactorily dispose of a combination of ≤3 liters of human urine and ≥400 grams of human feces for a total unit marginal cost of less than 5 cents. For economic estimation purposes, the cost of money should be assumed to be 6%/year, and facility operational lifetimes should be taken to be no greater than 20 years. Plausible economies-of-scale may be invoked in making this economic feasibility case.

So I think the one-pager must include the equation to convert capital cost into annual cost.
I was digging around and found this one which as far as I know is correct:



From the statement above, we know that the life-time for the RTTC technologies should be 20 years. So that is n = 20 years in the equation above.

The small i in the equation above should be 6% according to the RTTC conditions.
Only one small problem for me here: everyone calls the i something different.

In the statement above it is said "cost of money".
Maybe they meant rather “cost of capital”: “The cost of capital is the rate of return that capital could be expected to earn in an alternative investment of equivalent risk.” en.wikipedia.org/wiki/Cost_of_capital . This means it allows an investor to compare one investment option with another.

Christoph Platzer sent me these points by e-mail:
+++++++++++++
The use of the discount rate depends on the point of view intended.
If you look from a "volkswirtschaftlichem" (= national economy) view point, you have to use real interest rates. Real interest = normal interest - inflation. When you have a 8% interest and 6% inflation = 2% real interest.
If you look from a business calc stand point than you have to apply 6% or more in Brazil or less in Germany.

Considering the correctness...
a) the formula seams to be complicated but the application in an excelsheet is very easy as it is easy why use a simple but wrong calc.
b) Especially for long term aspects the application causes some difference.
c) the point you made "Annualised capital costs are easier for people to understand compared to NPV, even though with NPV you can be more accurate by taking into account in which year which replacement or emptying cost occurs (for example)." is the most important aspect to me.

++++++++

It leaves me still a bit confused, and I would probably just call it "the small i in the equation", and say that for the RTTC grants to compare them all it should be 6%; that's it.

Another aspect is that sometimes, for back-of-the-envelope calculations, O&M costs are assumed to be 10% of capital costs. Is this at all reasonable when no other estimate is available? Or is it only reasonable for conventional wastewater treatment plants or other more high-tech processes but not for pit latrines and UDDTs for example. (O&M = operation and maintenance)

Easy test: UDDT (urine-diverting dry toilet) has higher capital cost but lower O&M cost compared to pit latrine. So here in this example it is not right to apply the same percentage value to estimate the O&M cost from the capital cost.

Regards,
Elisabeth

P.S. By the way, Roshan from the BMGF told me that they think that 5 cent per person per day is still too expensive for the urban poor (this was also pointed out by Hanns-André in Sept. 2012, please scroll up or see here: forum.susana.org/forum/categories/139-ge...g-capital-costs#2226 ). Their future target may be more likely 2 cent per person per day. I think that's better.

,
I really don´t agree on using 10% of investment as O&M! This would be worse than no assumption. Why? It leads to wrong conclusions.
The operational costs are often as important for the total costs as the investment/capital costs. To express one as a percentage of the other therefore necessarily leads to wrong conclusions.

For instance the discussion about 2 - 6 - 10% discount is of much less importance than the O&M costs for the final result.

So please let’s not enter into such a simplified calc. Once we have exact data for one system maybe we could express for this specific system the O&M as a % of investment. Up to that point.... no way.

Regards
Christoph


P.S. I really appreciate the effort of Elisabeth to get this important discussion going as it is the base for comparison and sorry for the strong expressions, but I already internally wrote about the 10% aspect, which I consider very important (as you might have noticed

Oh, and of course the equation can also be inverted and used the other way around:
If I take the 5 cent per user per day as a starting value, then I can determine that the maximum price of a family toilet under the RTTC grant scheme can be:

490 USD

Assumptions of this calculation:

• n = 20 years in equation above
• i = 6% in equation above
• Number of household members = 5
• O&M costs = 10% of capital cost -- I know this is often not the case, but just for argument's sake.

My little Excel file for this calculation is attached below (tell me if you find a mistake in the calculation!).

If the criterion is lowered to 2 Cents per user per day, then the maximum allowable toilet capital cost (again for a family of 5) is:
195 USD

Still pretty unaffordable.
Anyway, one can now play around with the numbers and look at different scenarios. At least we can now see what 5 cent per user per day really means, and how much the costs still have to come down for some of the RTTC ideas if they are to go to the large scale and if their target group really is the urban poor. Could also be that 5 people sharing one toilet is not realistic then and one would have to go to 10 people sharing (community toilets). Etc.

Oh, I only now see the post of Christoph about the 10% of O&M. You are totally right. But it's just for argument's sake here for my inverted equation, or do you have a better assumption or way of calculation? I am talking super-simplified, just to understand what 5 cents per user per day really means...