Hi Kris, yes exactly. Sometimes the variables relating to each technology might be categorical (suitable for irrigating trees, suitable for irrigating crops for cooking, suitable for irrigating salad crops), other times ordinal (quality rating out of ten), or even numerical ($ value of costs). Everything relates to consumer utility. Consumer utility is preference represented as value, as judged subjectively by an individual in $ terms. The individual values a product by trading off benefits that offer a value advantage over alternative products, with the product’s overall costs. Start with the product (technology) and value all its part worths. Part worth values can be summed to calculate overall preference or utility.

cheers
Dean

I like the idea of an aggregate index value of some sort. So instead of "sludge" you would have "sludge 70" Vs. "sludge 20" and the lower (higher?) the value the closer the sludge is to a final end product (stabilized, mostly pathogen free sludge). A similar system could be set up for effluents.

I think that would also help with your algorithm as each technology could be specified as having certain input threshold values and modifiers to alter the index value of the output.

Dear Lukas, Dean, and all

Thanks for your answers: I think we finally all agree

Yes, there are two issues :1) the design specification of the biogas reactor, and in general the question of what is meant with « compatibility ».

And yes, I have given it a lot of thoughts and tried out several things … after all this is my thesis

The procedure I am suggesting based on my research so far very much goes along the lines you both described. I suggest three steps:

1. In the first step the potential technologies are looked at and hose, which are not feasible in the given context are screened out based on « hard facts » like slope, density, water availability, etc… (stakeholder-independent criteria).
2. In the second step, I suggest to generate all possible system configuration from the locally appropriate technologies and to select than a set of sanitation system options which is divers and of manageable size (and appropriate) and can feed into the decision-making process.
3. It is in the decision-making process where the subjective relevant criteria such as costs, effluent quality etc. are defined by the decision-makers (e.g. stakeholders. Investors) and looked at more in details in order to discuss trade-offs and select the preferred option.

An automated model like the one I have built is most useful if it is used as part such a facilitated structured decision-making process (e.g. as part of CLUES Step 5). It would by no mean ever replace a detailed feasibility study where systems are looked at more in detail and among other the issues « compatible, IF… » are looked at.

Having said this, I also believe that introducing more complexity into the compendium would be the wrong approach. The compendium can provide the input (the basis / the decision options), but the selection of the « most appropriate and most sustainable » sanitation system options will always remain a complex (and stakeholder and context dependent) decision making problem which cannot be fully generalized and automatized but needs involvement of all stakeholders, knowledge on the available options and their consequences, and a lot of discussions…

However, when I am talking about my work to others I often get criticized to generate system options that are not acceptable (e.g. septic tank effluent being used for irrigation) as people being confronted with the system options have a given context in mind and then intuitively directly further asses the « sustainability » of the suggested option. To discuss these issues with the SuSanA community helps to build a common understanding how these options I am generating based on the Compendium have to be interpreted and it helps me to better explain this to my public in the future – thank you very much!

I also often have been questioned what is really meant by the “biogas reactor” and why it is defined as it is. I actually also looked up in the old documents on the Eawag server last week and found out that in the « first compatibility matrix of technologies » (as part of the “WISDOM” project at the times of NETTSAF ), the only biogas technology was the UASB which is much more precise than just « biogas reactor ». However, as mentioned by you earlier, Lukas, the design of a « biogas reactor » is a continuum and it wouldn’t make things clearer to further define several types. So, I think I will stick in the future to the definition we have from the Compendium.

But… there is one more issue. The algorithm to generate valid system configurations from possible technologies based on in/output products is not an easy one. I have tested many ways of computing it and talked to many experts in graph theory, integrated modelling, network modelling, etc. at Eawag and at ETH and the sad conclusion is that for such a problem (where some loops are allowed and different streams are mixed, transformed and partitioned again) there is no easy implementable mathematical solution. Because of that, I had to implement a new specific algorithm. First, I computed a cyclic one which tested all possible combinations allowing for loops (e.g. a sludge of a waste stabilization pond going to a sedimentation pond and an effluent of a sedimentation pond going to a waste stabilization pond), but his algorithm required huge amount of computational energy and scaled very badly. I finally implemented an acyclic algorithm following step by the functional groups of the compendium and allowing for loops only in the functional group S and T because there we can assume that the technologies are located close to each other. This algorithm, however cannot deal with technologies that have the same output as input (i.e. sludge in the case of a biogas reactor) because it would generate endless chains of different variations of those. So, don’t be surprised if you will find two types of sludge in my final model but I will try to find a better name than « stabilized » sludge

Cheers, and thank you once again for your contributions!
Dorothee

Hi Lucas, I was thinking along similar lines - start with criteria like required effluent quality and required solids quality (these being variable and categorisable), then consider "sustainability" selection criteria like level of acceptable greenhouse gas emissions per user, level of discharge to waterways/water tables and use these to narrow the options. Then the options could be narrowed even further according to capital investment as annualised (including maintenance) cost per user less value of products that have value to the investor (like biogas, nutrient rich irrigation water with low levels of pathogens).

The results will be selected from all available technologies and combinations of technologies because the compatibilities of these is not in question, that is already known. The required answer is the technology or technologies that meet the needs of an "investor", someone who already knows their selection criteria but needs to monetise this.

Producing and utilising biogas does not influence the resulting sludge quality. It will be sludge (i.e. low quality solids because they require further treatment and therefore cost) whether or not you produce and use the biogas from the anaerobic digester. The solids being produced can be monetised as the cost of remediation, but producing biogas should win brownie points. A vermi-digester will produce a higher quality product (humus, requiring no further treatment). This product should earn brownie points, even though the digester cannot produce biogas. On the other hand, an anaerobic digester that doesn't capture and use the biogas should score lower because that system produces low quality solids and also discharges greenhouse gases.

cheers
Dean

Dear Dorothee and all

Looks like we are discussing two different issues here in this thread. Let me try to keep them separate.

1. Output products of biogas reactors / solids accumulating in digesters
I agree with Freya's concern about the solids: especially the fact that they occur in low quantities and frequencies increases the risk that they are neglected and not properly considered (e.g. the user may not be instructed what to do with it, or it may not be described in the manual of the plant). So it clearly has to be managed and shouldn't be ignored so it doesn't pose a risk.
I also agree with Kris who says that it is a maintenance issue, and the bigger concern is the safe day-to-day handling of the digestate.
Independent of the angle from which we look at it, and independent of the technology, it all comes down to proper O&M and management. This needs to be considered/planned for from the start of any project, and its programming ideally included as a project component (easier said than done in a context where projects of short duration are more common than long-term programmes). It sounds like in the Vietnam example described by Freya this was exactly the issue, and not the engineering design or the technology itself.

2. Compatibility of technologies - can this be described entirely based on input/output products?
I think this is the actual subject of Dorothee's point. Let me rephrase how I understand your problem Dorothee:
In your work you are going into much more detail than the Compendium, as for your detailed model you have to specify for each and every thinkable technology combination whether or not it is possible. You are developing an algorithm that allows to compute all the theoretically possible sanitation systems out of the different technologies available and compatible with each other. In other words, you need to know: can technology X be linked to technology Y? The computer needs to know "yes or no", while your answer may also be "Yes if", "No if", or "it depends". Let's take the example you brought: Can a septic tank be linked with irrigation, i.e. can the output product of a septic tank (effluent) be the input of irrigation?
Your problem here is that "effluent" or "sludge" (as used in the Compendium) are very broad terms, and not giving you the necessary resolution to answer this for each technology combination. Therefore, you are trying to increase the resolution by introducing more terms like "stabilised sludge", "treated effluent" etc.

This is how I understand your problem and how you are trying to tackle it. I would clearly say that it won't help to introduce more detail in the terms. It will still not be possible to cover all cases just based on the names of in/output products (there are always exceptions depending on the local conditions like tolerable risk, sludge/effluent quantities or characteristics, groundwater level, irrigated crop etc.). For example, if you want to irrigate a willow forest for biomass production in a sparsely populated low groundwater area it may be ok to use septic tank effluent directly, but not if you have salad on your field. Here, a more detailed description of the products alone may not be sufficient to describe compatibility.

Therefore, an idea I had is that you could possibly define your conditionality using other attributes/parameters (not just the name of the input/output products) to describe and model your possible technology combinations (systems). In that case you could stick to the coarse terms "effluent" and "sludge", as the necessary additional compatibility information would be stored in other, additional parameters that you would take into account.
I haven't properly reflected on such additional attributes, but they would probably be groupable as technology-dependent, design-dependent, context-dependent. They could go in the direction of "Water content", "Dilution", "Organic matter content", "Hygiene", "Purity" (i.e. absence of solid waste), "Level of stabilisation", or even "Risk of water resource contamination" etc. Some sort of grading (categories A, B, C, D, …) would then allow you to say for example irrigation of salad only by effluent with "Hygiene" level A, co-composting of sludge only if "Water content" below C and "Purity" above B etc.
Thereby, it needs to be considered that the grades of the attributes will not be entirely tech-specific (but also design or context specific). I am sure you have put much more thoughts into these aspects already…
Anyway, this approach would not really make your task easier - it remains very complex to define all these conditionalities (ideally based on scientific data - if available) for a universally applicable model.

The advantage of sticking to the "resolution" of input/output products as used in the Compendium is that it allows for a 2-stage approach:

1. With the simplified products "effluent" and "sludge" it allows technology combinations / system design based on what is "potentially compatible". Hence, it includes all (also questionable) system options/tech combinations, including e.g. septic tank effluent connected to irrigation, or pit latrine sludge directly connected to co-composting or land application. However, until here, subjectivity is basically excluded.


2. In a second stage, the decision-maker(s) can factor in context-specific information to validate systems and dismiss alternatives that are either not feasible, not recommended or not desired, based on different levels of judgement calls for the individual case.

Best,
Lukas

Hi again

According to the Compendium, the emptying frequency of a biogas digester is 5 to 10 years as opposed to a septic tank which is 2 to 5 years.
For the septic tank, the Compendium forsees two output products: effluent and sludge.
For the biogas digester, there is only sludge as an output product.
Based on the explanation of Lukas I do understand that the settled sludge in a biogas digester is not taken into consideration and that the "slurry" coming out of the biogas digester is rather a mix of effluent and partly digested sludge that, according to the Compendium definition of the term falls into the category "sludge".
This is something I can agree with a generic definition.
But I still have some problems with the very broad definition of "sludge" mainly because these products are also used to define the compatibility of two technologies. For example, I have difficulties to imagine that the treatment technologies that apply to the sludge of a septic tank are the same that apply to the slurry coming out of a biogas digester. The same thing applies to the sludge removed from a drying bed and the sludge removed from a waste stabilisation pond. This is the reason why I think we need some differentiation in the term "sludge" indicating its degree of dilution and its degree of digestion/stabilisation.
I would appreciate your opinion on this, especially the one from Lukas ?

Cheers, Dorothee

The settled solids in a biodigester that need infrequent emptying are in my opinion a maintenance issue. Just like you need to replace rubber seals etc. from time to time, it is necessary to remove these solids from time to time.

Of course this argument gets much less black/white when talking about fecal sludge from a septic tank, but even there the bigger day to day concern is the liquid effluent.

Dear all

Thanks to all of you for your valuable contributions and particularly to Lukas for these detailed and useful explanations!!!!

The question which we are discussing here has been popping up again and again in this work. The reasons is probably that there is no single answer to it (as demonstrated by this thread). And that is why it is worth to be discussed in detail publicly.

I fully understand (and agree to) all the explanations given by Lukas, but...
There are actually two main concerns I have about the current definitions in the Compendium:
1) "Settled sludge" versus "outflowing sludge"
The first one concerns the one mentioned by Freya. There is a settled sludge on the bottom which has to be removed more or less regularly depending on the design of the reactor and than there is the "slurry/digestate/...." that is coming out of the reactor continuously. As Freya mentioned, it should be clearly indicated that there are two projects. And these products are the most often also very different in terms of "their compatibility" with subsequent technologies. What brings me to the next concern...
2) Definition of products
Can the sludge that is removed from e.g. a "single pit" really be "applied on the field" in the same way that the sludge removed from a "drying bed"? Can the effluent from a "septic tank" be used in the same way for "irrigation" than the effluent of a e.g. "a waste stabilization pond"?
My answer is no. And this is why I introduced two different types of products for "sludge" and for "effluent". I thinks this is a feasible and valuable (at least for the sake of generating entire systems from the combination of compatible techs, and this is what I am currently working on) adjustment of the Compendium, given that these products are clearly defined (and do have an appropriate name*).

The question of what is a "valid" sanitation system is not an easy one and there is no single correct answer.
Why is it not OK to add pit humus to a biogas reactor but it is OK to directly use the septic tank effluent for irrigation? It all depends on the context and the expert opinion and thus the answer always involves some sort of subjectivity. This is why it is so important to discuss these questions and to try to find a common "best" answer to them...

Has anyone a good suggestion how to deal with the two concern above better than the way I did (a: having sludge and effluent as outputs for a biogas reactor, and b: introducing two types of sludge)?

Cheers, Dorothee

*) this is an issue I still need to solve...

Dear Dorothee, Lukas and all,
Thanks for the interesting discussion. I found Lukas’ synthesis a useful summary however am concerned that if there is no mention of the stored sludge just because its removal is irregular, as per the “not both” point, it may not be clear that there are two sludge by-products that require management. Particularly as I have seen that there was little consideration of the safe management of the liquid/effluent sludge in many biogas systems in rural Vietnam (discharged directly to waterway or reused without further treatment), and while there was awareness of the solid sludge there was no information on how to safely remove it once the systems fill (this is happening a lot in Vietnam as many systems pass 10 years).
Therefore while I agree with the definitions to call the liquid effluent “sludge” based on Lukas’ points, I think it would be beneficial to note that there are two sludge by products that require management – one that is discharged more continuously (assuming a water-based input such as a toilet) and one that requires infrequent emptying.
Anyway, happy to hear other views since I recognise there are many approaches to design and use of biogas digesters and maybe my comments are only relevant to the systems I’ve seen in SEAsia.

Kind Regards
Freya Mills

Senior Research Consultant
Institute for Sustainable Futures
University of Technology Sydney

Dear Dorothee, dear all

We had a very similar discussion at the time we were preparing the Compendium, so I am happy to contribute in detail.

First of all, I would like to list a few general considerations and clarify some confusion about the biogas technology:

• Digesters can have many different designs and applications, with toilet/wastewater inputs between 0% (e.g. agricultural or biowaste digester) and 100% (e.g. biogas settler in a DEWATS design). Whatever case, the technology can always be called biogas reactor, or digester. Defining 2-3 different technologies is not necessary, and in my opinion just confusing, as it is a "design continuum".


• In the context of sanitation systems (and the Compendium), toilet/wastewater inputs are always >0%. However, as pointed out by Kris, units with 100% wastewater input (sometimes called biogas settler, if installed as primary treatment unit of a treatment plant) are rare, due to the limited gas production which often doesn't justify a gas valorisation system. Hence, it is normally a mix of toilet/wastewater inputs and organic solid waste or cow dung, to increase the gas production.


• Whatever design or application case, there is only one (never zero or two!) main product coming out of the system: the digestate. Depending on the design, capacity and input, digestate can be:
  • between watery (very diluted) and rather viscous (depending on solids concentration). However, it is always a liquid "slurry", so it can flow through the reactor
  • generated in small quantities or large quantities. The amount of digestate coming out corresponds to the amount going in (communicating vessels). Hence, it is not correct to say that quantities are small or negligible!


• The expansion chamber does not primarily serve as a solid-liquid separation or settling chamber (although it could be designed for that, e.g. in a biogas settler to keep solids in the digester - see some BORDA designs with slanting chamber floors). Its function is to provide an "expansion" volume for the slurry that is pressed out by the gas pressure (if the gas is not being used). In this way, gas can be stored inside the digester, and digestate in the expansion chamber, until they are used. The expansion chamber as a storage volume is mainly important for flexibility in systems where the digestate is removed manually (and brought to the fields, for example), and not if you have a sewer pipe connected to transport it.

Now, regarding the input and output products used for the Compendium, here is the explanation:

• As already mentioned by Dorothee, the Compendium tries to reduce the complexity. It is not a design manual. Hence some simplifcation to focus on and facilitate the general understanding of key relevant aspects.


• The inputs and outputs shown for the different technologies should help to identify compatible technologies that can possibly be linked in a meaningful way. However, the appropriateness of these potential linkages is very context or design dependent. More design knowledge and experience is needed in a subsequent step to assess that.


• Effluent definition as per Compendium: "Effluent is the general term for a liquid that leaves a technology, typically after Blackwater or Sludge has undergone solids separation or some other type of treatment. Effluent originates at either a Collection and Storage or a (Semi-) Centralized Treatment technology. Depending on the type of treatment, the Effluent may be completely sanitized or may require further treatment before it can be used or disposed of."


• Sludge definition as per Compendium: "Sludge is a mixture of solids and liquids, containing mostly Excreta and water, in combination with sand, grit, metals, trash and/or various chemical compounds. A distinction can be made between faecal Sludge and wastewater Sludge. Faecal Sludge comes from onsite sanitation technologies, i.e., it has not been transported through a sewer. It can be raw or partially digested, a slurry or semisolid, and results from the Collection and Storage/Treatment of Excreta or Blackwater, with or without Greywater. Wastewater Sludge (also referred to as sewage Sludge) is Sludge that originates from sewer-based wastewater collection and (Semi-) Centralized Treatment processes."


• From these two terms ("Products" as per Compendium terminology), it is clear that the digestate can be classified within them (no need for and no added value from having it as a separate product - avoiding unnecessary complexity)


• However, we had long and detailed discussions about the outputs of the digester. Is it effluent, sludge or both?
  • Not both: as described above, there is only one main product leaving a biogas reactor: the digestate. It may be required to empty a digester from time to time (e.g. every 10 years), to remove the solids accumulating at the bottom and reducing the effective HRT. With the anaerobic processes and rising gas bubbles in the reactor, solids are kept in suspension and the slurry is mixed, but heavy particles (grit and sand) or floating materials (e.g. straw) may accumulate with time. As both the amount and frequency for removal of such solids are normally very low, they are not highlighted as a separate output product in the Compendium.
  • Effluent or sludge? Two main considerations: 1) Digestate normally has a high organic and solids content (consequence of using input substrates with high biogas yield), so it is closer to a sludge than to an effluent as per the definition above. 2) If you look at the downstream treatment requirements for digestate, you see that they commonly correspond to those of sludge (e.g. you wouldn't put digestate into a constructed wetland, as TSS is too high) - as already mentioned by Dorothee. Further treatment of digestate may be needed, but not always (depending on substrate, digester design and end-use).

Let me summarise the terms used in the Compendium for the inputs and outputs:

Inputs:

• Sludge
• Blackwater
• Brownwater
• Organics

Outputs:

• Sludge (see explanation above)
• Biogas

And here the terms not used, with justification why I do not consider them as appropriate:

Inputs not used:

• Pit humus: earth-like solid, possibly similar to compost, possibly with more inert fractions. Not suitable for digestion. Can directly be used as soil conditioner.
• Stored faeces: already dehydrated, often mixed with problematic bulking material. Generally unsuitable for digestion. While theoretically possible, there is no point of dehydrating faeces before wetting them again and digesting them, so in practice a negligible option.

Outputs not used:

• Effluent: see explanation above.
• Digested sludge: unnecessary complication to call it digested sludge. More precise to call it digestate directly.
• Stabilised sludge: you define this as "sludge that has undergone some sort of treatment": this is neither precise, nor helpful, as treatment processes are diverse and can include solid-liquid separation, stabilisation, dewatering, and pathogen reduction. Hence, a product can have undergone some sort of treatment without having been stabilised. "Can be reused directly" in the fields is also not a good definition of stabilised sludge, and not precise for digestate. As mentioned earlier, the level of treatment in a digester and the requirement of further treatment depend on multiple factors. So this term wouldn't generalise it well enough.
• Treated sludge: this may imply the message that it is a (safe) end product, which is not always the case (in fact, pathogen reduction is not a strength of mesophilic digesters).
• Secondary sludge: this commonly refers to sludge removed from secondary clarifiers in wastewater treatment plants (as opposed to primary sludge from primary sedimentation tanks).

For more information on the Compendium logic, please consider the following sections of the Compendium:

• Biogas Reactor factsheet (p. 80 and p. 134)
• Treatment overview (p. 98)
• System 5: Biogas system (p. 28)

Hope this helps!
Best,

Lukas

dorothee.spuhler wrote: Hi again Kris!
Re "about the stabilized sludge...": what about "secondary sludge"?

That would be also confusing as "secondary sludge" most commonly is used for excess sludge from an aerobic activated sludge process and is in fact often an input into a biodigester. I guess "digested sludge" would be the least confusing?

dorothee.spuhler wrote: But my initial question remains
For (2) and (3): do we need to consider them having an effluent or due with assume it is either not significant or mixed with the "secondary sludge"?

I gave it some more thought, and I would say no effluent. Not so much because there isn't a more or less liquid waste product but rather that contrary to common septic tanks etc. that contain huge amounts of household grey water (easily 95% of the waste volume created), the sludge is the dominating waste product for a biodigester. In addition there is the regulatory dimension with "effluent standards" etc. none of which really apply to waste products of a biodigester as it will be nowhere near a level that would allow discharge into a surface water body as by regulations.

Hi again Kris!
Re "about the stabilized sludge...": what about "secondary sludge"?
Re "the thee technologies": totally agree, (1) is covered by "settler", "septic tank", etc..

But my initial question remains
For (2) and (3): do we need to consider them having an effluent or due with assume it is either not significant or mixed with the "secondary sludge"?

Cheers, Dorothee