MULTIPLE ON SITE DETENTION

Unified Conveyance–Storage System in MiTS 3 StormSync #

Traditional design calculations focus only on the conveyance system under steady flow conditions, assuming constant velocity and discharge over time. While this makes calculations simpler, it does not capture the real dynamic interaction between conveyance and storage during storm events. In drainage systems with interconnected ponds, detention drains or OSDs, the flow changes over time, making steady-state analysis no longer optimum or valid. MiTS 3 can now integrate the conveyance and storage unit into one continuous system where inflow, outflow, and water levels are calculated at every time step.

Multiple OSD connected to Main Drain (Parallel Layout) #

In this section, we will discuss the hydraulic behaviour when multiple OSDs are connected to the main drain in parallel arrangement (assuming that each housing will have underground OSD). The objective of this study is to compare the hydraulic results using MiTS 2 Steady State Method, MSMA 2nd Edition (Table Method – Chapter 5) and MiTS 3 Dynamic Wave Routing. 

The drainage layout follows the sequence as: Individual Drain > OSD Tank > Outlet Drain > Main Drain as per image below.  

Sample file

MITS 3: Dynamic Routing

MSMA 2nd Edition: OSD Design

Information

Result comparison #

Steady-State Method and MSMA Table Method #

Table 1: MiTS Steady-State Method vs MSMA 2 Table Method

Note that MSMA table method is effectively a conservative approach by making the storage size very big in order to reduce the discharge from the OSD. So the table gives a smaller inlet flow and outlet flow. That can make them smaller than a direct analytical rational result (CiA/360).

From the result above, the post development outflow calculated using the rational method (CiA/360) is larger than MSMA 2.0. Engineers usually design their drainage system by:

1. Rule of thumb approach,

Where the larger peak flow will be adopted for drainage design, on the assumption that there are no storage units inside the drain network at all. However, under this assumption, it is uncertain whether the post-development outflow (Qpost) will be lower than the pre-development outflow (Qpre).

2. Total up the PSDs from each OSD as the Qpost

If multiple OSDs are connected to the same main drain, their outflows (PSDs) are simply added together to represent the total post-development discharge (Qpost) which should then be compared against the pre-development flow (Qpre) for compliance checking. But then we are still unsure if the Qpost will be smaller than the Qpre or not. 

Both approaches share a common limitation, as they rely on simplified assumptions that do not accurately capture the dynamic interactions between storage and flow routing within the drainage network.

Dynamic Wave Routing Analysis #

 Now, we will incorporate the conveyance system from the drainage design and storage system from the OSD MSMA design in MiTS 3 Analysis Mode. Based on MSMA 2.0 table method, the suitable OSD tank size based on SSR is 8.0m x 5.0m x 1.5m (Vol,prov = 60.0m3) with orifice diameter of 70mm as the primary outlet. 

To evaluate the hydraulic performance, two alternative OSD01 tank sizes with different storage volumes were analysed under the two orifice diameters (70 mm and 100 mm), to observe their respective inflow–outflow hydrographs.

 Orifice diameter: 70mm

Table 2: Outflow and Water Depth value of MSMA size vs alternative size in Dynamic Routing

Orifice diameter: 100mm

Table 3: Outflow and Water Depth value of MSMA size vs alternative size in Dynamic Routing

For a single OSD (OSD01), although the MSMA 2.0 table method recommends a larger storage tank with a 70 mm orifice, the analysis shows that even with smaller alternative tank sizes and a larger 100 mm orifice, the post-development outflow (Qpost) remains lower than the pre-development outflow (Qpre). With dynamic wave routing, smaller storage designs are still capable of achieving sufficient attenuation performance while offering a more space- and cost-efficient solution.

 Discussion #

1. Why is the outflow different though the orifice diameter is the same for varying storage sizes?

From table 2, observe that the outflow increases when the storage size decreases despite having the same orifice diameter. This is because the storage size indirectly influences how the water depth/ water head builds up above the orifice, which will then affect the outflow discharge.

By orifice equation;

Q=Av=C_dA\sqrt{2gh}

 Since the orifice area (A) is fixed, the outflow rate (Q) is proportional to water head (h). For a larger OSD size, it provides larger storage volume. This large volume can temporarily store the runoff (inflow) without the water level rising significantly, which reduces the hydraulic head and reduce the outflow rate

Meanwhile for smaller OSD size, when the same peak inflow enters this smaller tank, the water level will rise higher to store the volume temporarily, creating a higher head pressure on the orifice, thus resulting in higher peak outflow discharge at the orifice.

2. Storage utilization

Based on the analysis, it can be evident that the storage tanks designed using the MSMA manual are considerably larger compared to those optimized under dynamic routing simulation. The results show that the maximum water depth in the MSMA-based storage remains relatively low throughout the event, indicating that a large portion of the available volume is not effectively utilized. This underutilization demonstrates that the conventional static method (MSMA) tends to produce conservative designs with excessive storage capacity.

In contrast, the dynamic routing approach allows for a more realistic simulation of inflow/outflow interaction, accounting for flow attenuation and hydraulic control effects within the network. As a result, it achieves higher storage utilization with smaller tank volumes, reflecting a more efficient and cost-effective design without compromising hydraulic performance. This leads to a notable point of discussion with the approving authority, where engineers may justify that the use of smaller OSD tanks can still achieve compliance with the regulatory intent while offering a more cost-effective and practical design solution.

3. Peak flow attenuation reduction (Qpost < Qpre)

The post development outflow obtained from PSD MSMA Table Method and Dynamic Routing will be compared with the pre-development outflow.

1. Single OSD (OSD01)

 Table 4: Orifice discharge comparison on different OSD storage sizes

2. Multiple OSDs (OSD01, OSD02, and OSD03) discharge to the Main Drain

The storage sizes below (column 3 and 4) will be used for comparison for Qpost vs Qpre in Table 6.

Table 5: OSD sizes in accordance with MSMA 2.0 and alternative size in MiTS 3 Dynamic Routing

Table 6: Post-dev and Peak-dev outflow comparison on different OSD storage sizes

From the inflow/outflow hydrograph, it can be observed that the post-development peak flow (Qpost) is lower than the pre-development peak flow (Qpre), even when multiple OSDs are integrated within the network. This clearly demonstrates that the inclusion of storage systems within the drainage network effectively attenuates the peak discharge before it reaches the main outlet.

Though the permissible site discharge (Qpost) for each of the OSDs from the MSMA 2.0 table method is way more lower than Qpre (difference of 46.43%), this is overly conservative, since it is on the assumption that all of the peak flows from the OSD happen at the same time. In reality, such excessive conservatism can lead to oversized storage tanks and unnecessary construction costs. 

With dynamic routing, it allows the system to capture the changing inflow and outflow between all connected OSDs. This shows that the system can redistribute and delay runoff during peak rainfall, resulting in a broader hydrograph with a lower peak

4. Self Inconsistency within MSMA table guideline

The recommended storage size and orifice diameter for all OSDs will be summarized in table below:

You can fact check the Qorifice vs PSD value using example in MSMA 2nd Edition Chapter 5 and input the information in the MiTS MSMA Module. 

Following the MSMA 2.0 table method, the guideline provides the permissible site discharge (PSD) and recommends an orifice diameter based on the lesser value from Table 5.A3 or Table 5.A4. However, when the recommended orifice diameter is used in the orifice equation (Equation 2.6), the calculated discharge (Qorifice) exceeds the PSD. 

This simply means that the MSMA recommended value of orifice size will have more than Permissible Site Discharge (PSD), supposedly the lower bound of the discharge to the main drain. So as an engineer, either you follow the given Orifice size or the given PSD, you can’t fulfill both conditions at the same time. 

On the other hand, if you follow the routing method, then you will get consistency because everything flows from the first principles of equations rather than some tables with magic values. 

But then, it’s not our position to say that the MSMA table method is wrong or what. We are here just to point out the inconsistency so that the engineers can be better informed.

Conclusion #

1.  In MSMA 2 Table method, the storage unit is made very big, so that the Qpost ( PSD) is made very small, in order to satisfy the ( written or unwritten) requirements that Qpost < Qpre. 
2. The Table method is further overestimating the discharge to the main drain if we have parallel OSD, because of the time lag effect
3. An additional problem that we discover with the Table method is that the recommended Orifice size and the PSD can be inconsistent with each other. So you have to pick. The dynamic routing doesn’t have this problem. 
4. Dynamic routing allows us to have more parameters to play with ( eg: varying the size of storage units and orifice), within the existing design constraints such as Qpost<Qpre. It does  this by having an actual simulation based on the actual hydrograph of the site.

Speculation #

Based on Conclusion 1, we speculate that the MSMA 2 designers do a lot of dynamic routing simulations in arriving at the table values for sizes, PSDs, inflows and so on. So it’s hard work and the focus is more on not flooding rather than economical design.

Despite that, there is simply no way that you can just capture the complexity of hydraulic modelling within a single table with values, and hence the PSD/Orifice size inconsistency we observe above.